Taking this off in a different direction.
The selection pressure driving these things into existence is HIGHLY convergent all around the world, with about four mutations appearing again and again and again. This is adaptation to the human host, not really evolution for immune evasion, especially considering that only like 30% of most populations at most has been infected.
You would expect immune evasion selection to be wildly divergent rather than convergent. What is mostly happening is an interaction between the fact that regions of spike protein that are on the surface of the spike and are involved in host interactions are evolving to do do well in a new host, and the fact that these functional regions of host interaction are under many selective pressures and more likely to be 'sticky' and places where functionally important antibodies tend to bind. Slight immune evasion is a side effect, not what drove them into existence in the first place, for the most part.