One example of an alternate history that came to mind is if Maynard managed to discover his sieve a few years earlier, then Yitang Zhang may have been discouraged from his bounded gaps in primes project, and the ensuring Polymath project would never have gotten off the ground. In many ways, the ordering of results was actually quite fortuitious.
In additive combinatorics, I sometimes wonder what would have happened if the striking applications of the polynomial method and the entropy method had come much sooner, so much so that Fourier methods would have been viewed as uncompetitive and studied with far less intensity. The subject of higher order Fourier analysis, in particular, would likely have taken a far different path...