This notion appears to go against all the expert advice against (ab)using anti-biotics. Sure, less infections means less chance for mutations, but due to the fundamental forces at play, the resistant mutations will spread faster through the resistant population and eventually become the prevalent strain.

Unless you can eradicate the virus or bacteria completely (mass effective vacination), the tools we create to fight them are really only delaying the inevitable.

>If some variants do evolve that escape the vaccine, odds are good that they will be less dangerous than the original, not more.

From my understanding, this is not the case. It's pure chance whether a mutation is advantageous or disadvantageous. If you are looking on a per-mutation basis, then sure, it's likely the mutation is disadvantageous. But at a population level, it's likely that atleast one advantageous mutation will emerge reasonably quickly when we are talking about something at the scale of several thousands of infections.