Another observation of the SE process over the years to add here is that the proposals are somewhat a “snapshot” of the given moment in time at which the proposal was done.
We generally don’t go back to reword or adjust proposals if say “Swift N+1” lifts some restriction that a past proposal might have referred to as “impossible” etc. So there’s few reasons to do amendments other than something which may have been incorrect in the proposal text (like a missed final change when a proposal is “accepted with changes”, or perhaps some section was left in there by accident, or just not implemented entirely etc).
That’s good, since they’re all just incremental and documenting progress of the language. It’s also how the proposals sometimes can be insufficient as “documentation” and live documents like the swift book still are super valuable, as those should be continuously updated to reflect the current state of the world 