Before we get too far into this can we stipulate that the answers bet somewhat realistic? I know there are going to be people that post their dream list of changes...$100 billion of changes that would work but has no chance of ever happening - even if John Lasseter became CEO.
So that being said some of the easiest things that could be done would be to limit attendance or raise ticket prices further. And do it by a large degree. If attendance at the MK was limited to 20,000 or each ticket cost $250 a day, you'd get significantly reduced crowds. However, neither of these are great ideas as they wouldn't do anything favorable to Disney's reputation and I personally would hate them. So that's out.
Indeed expanding the current infrastructure as they are doing would help. Wider pathways, more transportation, main street bypass etc. This is probably the most realistic easiest thing to do, and I'd expect to see more of it. You optimize the resources you have.
Another thing that you'd want to do is to spread out your crowds more evenly. Within the park as well as throughout the resort. This is a terribly tricky thing to do though. The fact is certain rides, and indeed certain parks are more popular than others and you can't just ignore that fact. You can build all the attractions in the world at the other three parks, hell build another 3 parks, but nothing is going to change the fact that MK remains the most popular, and that nearly everyone that goes to WDW is going to want to visit it. All you accomplish for the MK by building attractions or parks elsewhere is increasing the attendance at the MK and at least in the more park scenario, likely decreasing attendance at AK, and DHS. And unfortunately the biggest capacity problems are seen at the MK. However there are capacity problems at the other parks as well and I think this means that the problem would need to be addressed in different ways for the different areas of the resort.
For AK, and DHS, the capacity problems largely manifest themselves in long attractions wait times, particularly at DHS. This is simply because there are too few attractions for the amount of guests. So the strategy outlined above would actually work fairly well. I don't think too many guests are attached to visiting DHS or AK on their vacation so building more parks would likely decrease attendance at each. Interestingly, letting Universal get more popular would also achieve this end, without incurring the expense of an additional park (which the resort doesn't need anyway). Of course, this strategy still hinges on a decrease in attendance at the these two parks. Probably not something that WDW would want. And because there are so few rides, and most are headliners, the strategy I'm going to outline for MK, wouldn't work as much here. The best option for AK, and DHS is to simply increase the count of attractions, as well as shops, and restaurants - in particular large people eating attractions would be best. Most people would probably like to see some sort of e-ticket ride, but you would need several to really effectively address the problem. A large capacity show ala LMA, or Indy would also do wonders, but it would have to be quite popular and have many shows a day to have lasting impact. The biggest issue with these plans is the cost and practicality. How do you build multiple large attractions quickly without spending huge sums of money while simultaneously not trying to increase your attendance too much? My guess is that you could pick any two of them. It's not idea but if you slowly build up the amount of high-capacity attractions in the park you would be able to grow your capacity without drawing the massive crowds that a massive expansion would attract.
Epcot and especially MK are harder. Those that suggest adding multiple smaller rides are probably on to something. Unfortunately the kind of rides that work best elsewhere I'm not sure would work here. Disneyland has half a dozen classic dark rides. While it wouldn't be terribly expensive to do this at the MK (assuming you could find the space), I fear that given the type of guests WDW receives this could severely backfire for a number of years until the rides finally lost their novelty. The prime example of this is LM, it is not all that great of a ride, yet can command 2 hour waits because people that come once in a lifetime want to experience everything. The same ride at CA is a walk on. Now while some of this effect is due to FP+ I don't think anyone would seriously attribute more than half an hour of the wait to it. So if a relatively mediocre ride with a 2000+ person/hour capacity can generate 2 hour waits would a slew of brand new, much smaller rides work as intended to fix the capacity problem? A different problem is generated if you build new large e-tickets, as they'll increase crowds. At the end of the day, new attractions have to be a part of the plan, but the only remedy I can think of to offset the "new attraction" guest craze is to build them slowly one at a time. This also is the only way the company could afford to do it. Ideally, you'd want to nearly double the size of the kingdom but keep the attendance relatively stable. The cost of this would be enormous and is only practical when spread out over a long time. The problem of course, is that while the capacity issues would eventually be resolved - little to nothing ends up being done now where the problem actually is.
This then circles back to issue of distribution of the crowds. Ideally you'd have some way to have an even amount of guests entering the park every hour, as well as each attraction - rather than having huge fluxes of people swarming towards certain parts of the park at certain times of the day as is now. If you could nudge your guests in the right direction by say, telling a couple thousand to go to tomorrowland right as the parade starts you could ease congestion on main street and frontierland. Heck if you could even find a way to incentivevise them to watch from frontierland rather than the hub you'd be 5 steps ahead. Better yet, you'd want a way to be able to analyze where all the crowds are currently and how they've acted in the past so you'd be able to accurately predict where crowds were going to be ahead of time and try to compensate. This system wouldn't solve your capacity issues. Ultimately the capacity of the park is fixed until new attractions are added. But if you could control all your guests touring plans then theoretically you'd be able to make it so that they'd all be evenly dispersed throughout the park - making the park feel much less crowded. Of course most guests would be rightly angered with Disney telling them where and when to go in the resort or the park. So you'd have to find a way to get guests to agree to it while thinking they were still in charge. For instance you could do it under the guise of helpful suggestions. Ask the guest what attractions they'd like to see most, determine what areas of the park those attractions are, and (under the assumption that most guests ride adjacent rides before moving on to further rides) suggest to the guest the "best" times they could go to those rides to expect the least wait. And of course Disney would naturally be suggesting different times to different people to maximize the spread of the crowd. They could even go so far as to offer an incentive to follow through on these suggestions - perhaps by offering a reserved spot on that attraction at the suggested time so that the guest would have a guaranteed short wait. Thus the guest is in the desired land at the desired time and crowds could be better managed. The system could be easily expanded to function on the resort as a whole rather than just on a park basis.
Yes I'm describing FP+ - if implemented correctly it can be a giant step in the right direction. The genius of the system is the means by which the FP's are suggested. You don't initially pick the time - just the experiences. Then the system automatically picks times for you. You're free to change them, but even in my case as a well experienced tourer, I seldom have because they're more or less what I wanted to begin with. So really what is going on is Disney has persuaded me to go along with their touring plan that undoubtedly is being used to spread crowds evenly throughout the park. FP+ doesn't have the drawbacks of drawing a ton more people to the park, or drawing a bunch of people to a specific park or corner of the park. But having a system like it makes so much sense, its a wonder that its only now being implemented. Rides and attractions are essential to the the long term capacity needs of the resort and all the parks, but for both practical, and financial reasons really must be built at a fairly slow pace - especially if there is no system in place to manage the crowds they would generate if built faster.
So those are essentially the ways to do it that I can think of. One think I didn't pay too much attention too and has been slowly disappearing from WDW are more restaurants and, in particular, shows. There was a time when there were several more stage shows throughout the MK. And while I'm actually more of fan of the Disneyland "no stage shows" approach, opting instead for small-scale entertainment, there is no denying that a 7 time daily stage show can easily draw 1000, maybe even 2000 people away from the rest of the park if given the proper space. It'd also be fairly inexpensive - even at a high level of quality with live music. Oh how I miss a live pit for the Galaxy Palace theatre and castle stage shows. If the aim was specifically capacity based though, you'd want to stay away from any more castle/main street entertainment and move to the other lands. A frequent live stage show in both Adventureland and Tomorrowland could take a combined 4000 (or more if people are waiting in line for the next one) people off the streets and rides. The same would work for Epcot, and in its case several of the venues already exist or could be repurposed. Even better, the cost is nowhere near as high as an attraction, shows are easily swapped out, and attendance increases would most likely only be modest.
Alright, I'm out of ideas for now. Thanks for an interesting thread.