Another example. In medicine, we often need to make initial treatments choices before we have all of our necessary clinical data. As this comes in, you fine-tune the plan, or in some cases, you make a complete 180 degree turn in treatment strategy when the initial suspicion ends of up being wrong. You don't just observe the patient and possibly let them crash while waiting for result of every test. For example, most patient's will receive initial empiric treatment for pneumonia based on what antibiotics will likely work, not based on precise culture results that demonstrate the exact infecting organism and its antibiotic sensitivities. In most cases, this information will never be known, but the "good enough" treatment will work nonetheless.
We can extrapolate this idea to public health. Right now, I don't think we need to rush into doing anything different until we either know more about omicron or have evidence that out existing strategies for dealing with COVID will be inadequate for this variant (but I do think we should be doing more against delta anyway, but that's a different debate...) However, we should take the necessary "do no harm" precautions, like at least doing the preliminary work on reconfiguring the mRNA vaccines, in case this ends up being needed.