Yes.
I'd prefer not to, but maybe you could find a way to classify gaps based on their strength and 'where' the day opens? If you spend a few weeks thinking on it I'm sure you'll discover something.
Some quick numbers.
If we look at all + gaps without any further filtering, 49,2 % close fully.
If we're less strict and accept within 2 points as a fill, we get 57,7 % of + gaps closing.
To the downside, the number is 52,4 % for a full close. Reflecting the bullish nature of the markets.
From these numbers alone it's clear that merely fading a gap is a stupid idea.
You could just simplify all this and merely pay attention to the prior Close. If we're moving towards it, generally, it's a good target. But if we're moving away from it - well...
I know some use the 16:15 close, but regular trading hours for US Equities are 09:30-16:00. So I've always used that one.