The theory that it is unusually difficult to sweep a doubleheader is actually proven to be false.
https://www.beyondtheboxscore.com/2017/7/19/15982598/doubleheader-sweep-probability-analysis
Theoretically without any other variables the odds of sweeping should be 50%, the data from 2008-2017 suggests that it’s actually slightly favored towards being more likely to sweep than to split.