Victor, can you demonstrate any cases of real code where this optimization makes a significant difference?

There are many, many tiny optimisations we *could* be making in Objects/longobject.c; each of those potential optimisations adds to the cost of maintaining the code, detracts from readability, and potentially even slows down the common cases fractionally.  In general, I think we should only be applying this sort of optimization when there's a clear benefit to real-world code.  I don't think this one crosses that line.

In the (I suspect rare) cases where a piece of real-world code is slowed down significantly due to a non-optimized 2**n, the code author still has the option of replacing that piece of code with 1<<n manually.  And in some cases, that's probably the wrong optimization anyway: an expression like `x * 2**n` would be better hand-optimized to `x << n`.

IOW, I'm -1 on making this change.