In this case my part of the sprint probably would have run over by one day, since I had one more feature to implement around the new core I had written and I had to spend extra time refactoring code for existing features a little (not too badly) to make them work with said new core code.
The logical thing would have been to punt that one feature, make everything else work (it was very close), then spend the rest of the sprint testing before swapping the old boat anchor out. Either that or keep the old version in place until the end of the next sprint to give time for complete unit tests to be written.
Apparently the fact that I had "rewritten something for performance" was viewed as an inherently bad thing to do regardless of the rationale or benefit. That's because some smart guy once said it was "the root of all evil," and everyone believes smart guys who are quoted a lot.
I really have a pet peeve about cargo cult thinking in general. Concepts and ideas have context. To me it's a sure sign that I'm dealing with an impostor who does not actually understand their subject. Yet another reason losing that job was a blessing in disguise.
The logical thing would have been to punt that one feature, make everything else work (it was very close), then spend the rest of the sprint testing before swapping the old boat anchor out. Either that or keep the old version in place until the end of the next sprint to give time for complete unit tests to be written.