This is the part I'm having trouble understanding. 10% doesn't seem like a high percentage for a distributor, even if they're just hosting text and images. Of course Substack also provides comments, notifications, admin/moderation, and some degree of marketing and aggregation effects (not Aggregation in the strict Ben Thompson sense, just people going to a well-known website because they know there are things to read there).
None of that is super hard to set up, but If I was Matthew Yglesias, and I was making $775k next year on Substack, how much incentive would I have to leave and pay someone to create and/or manage a new site for me? Even assuming there's no lock-in whatsoever from Substack, or that Substack does not increase its value proposition for content creators by then, it seems like I'd probably just stay where I was and collect my $700k, and be pretty thrilled about it.
It's possible I'm wrong about that in some cases, but would it happen in so many cases that it poses an existential risk to Substack? Seems like a stretch!