> I feel like there must at least be anecdotal stories of people reading, exploring and making connections they otherwise would not have been able to because free > $35/paper.

And I wonder how many of those people don't know about DeepDyve.com? It's essentially Spotify for journals. The $49/month ($30/month if you buy a yearly subscription) plan gives unlimited online access to most papers from a huge list of journals [1] from a huge list of publishers [2]. (There are also group plans available, but they don't list the prices for those on their website, as far as I can see).

You can also buy 5 packs of rental tokens for $20. A rental token gives you access to an article for 30 days. People on the free plan get 5-minute full text previews, so you can at least skim an article to get a better idea if you want to consume a token to rent it.

Open access is much nicer, of course, but most cases I've seen where someone says that the cost of non-open access, such as the $35/paper so many publishers charge if you buy directly through them, was a serious issue to their research would have been able to get those papers through DeepDyve at a cost well within their budget, but for some reason very few people in the wild seem to have heard of it even though it has been around for quite a while.

[1] https://www.deepdyve.com/browse/journals?page=all

[2] https://www.deepdyve.com/browse/publishers