> Is the position of that window relevant?

Who knows!

This seems like one of the things that makes it hard to test intermittent fasting: there are a bunch of variables other than just the length of the window.