There's a sort of joke in academia, about the "least publishable unit" for a paper: Since "how many papers produced" is much easier to measure than "impact on a field", one way to try to game the system is to figure out the smallest amount of research that can be called "one paper", and publish just that, to maximize the papers / research effort ratio.
What we're seeing now is the same thing for streaming services. Sure, you'd pay £20 for a subscription to watch 75% of all available content. But it turns out most people would pay £40 for two subscriptions, each of which would show you 35% [EDIT and even people who won't, will still pay one £20 subscription for 35% of the content]; and quite a few people would pay £100 for five subscriptions, each of which will show you 12%. The beancounters are busy experimenting to find the "least bundle-able unit", to maximize extraction.
EDIT: And as someone else has pointed out, this is not something that any Netflix -- or any potential replacement -- can unilaterally do something about. In fact it's a "tragedy of the commons" situation: If MGM and Universal Studios and Paramount and WB all license their content to Netflix (under the £20 umbrella), and Disney doesn't, then Disney gets to keep a massive amount of money for themselves, while the other rights-owners have to divide up the rest. It only takes one or two "defectors" to basically force everyone to do the same thing.