Paid-only Substack posts get you money from people who are willing to pay for the posts, but reduce both (a) views on the paid posts themselves and (b) related subscriber growth (which could in theory drive longer-term profit).
So if two strategies are
entice users with free posts but keep the best posts behind a paywall
make the best posts free but put the worst posts behind the paywall
then regarding (b) above. the second strategy has less risk of prematurely stunting subscriber growth, since the best posts are still free. Regarding (a), it's much less bad to lose view counts on your worst posts.
Not sure if you intended this precise angle on it, but laying it out explicitly: If you compare a paid subscriber vs other readers, the former seems more likely to share your values and such, as well as have a higher prior probability on a thing you said being a good thing, and therefore less likely to e.g. take a sentence out of context, interpret it uncharitably, and spread outrage-bait. So posts with higher risk of negative interpretations are better fits for the paying audience.
4Dagon
Substack started off so transparent and data-oriented. It's sad that they don't publish stats on various theories and their impact. Presumably you don't have to be that legible with your readers/subscribers, and you can test out (probably on a monthly or quarterly basis, not post-by-post) what attributes of a post advise toward being public, and what attributes lead to a private post. The feedback loop is distant enough that it's not a simple classifier.
You're missing at least one strategy - paid for frequent short-term takes, free for delayed summaries.