It's an easy target that neutralizes point of entry, but that also adds a great deal of selective pressure for mutations of the spike proteins, which we already

It's an easy target that neutralizes point of entry, but that also adds a great deal of selective pressure for mutations of the spike proteins, which we already know have an effective/successful mutation rate much higher than the rest of the virus. (Likely for the same reasons.) But the ACE2 receptor is relatively lenient in what it will pair with compared to the immune system. So a spike with multiple mutations is more likely to be able to continue infecting cells via ACE2, than the mrna-vaccine-induced-immuno-response (targeting only the spike region) is able to block it. And because of this "fuzzy" matching, less than perfect mutations can perpetuate until other helpful mutations occur. But each of these mutations is also a step closer to evading detection by the immune system, so they can be helpful over multiple iterations of mutations, even if there's no selective pressure/benefit provided per mutation. That it can continue to bind to ACE2 is enough, until it reaches a point where the immune system doesn't recognize and must re-learn that target.