From what I've seen of the people in charge of Signal- they'd probably close before they sell out.
That said, you make a very good point. Having all the registered users in one place, is a vulnerability. A great many of us have non technical friends/partners/siblings/coworkers/etc; and encouraging them to use ANYTHING new is pulling teeth. So Signal is great, but it's still eggs in one basket- if they do something user-unfriendly or sell out or close, we are back to square one in begging/pleading/cajoling people to (please) try this (much better) app.
I've also lost a few people who used Signal over one stupid problem- the iOS version has no backup/restore function. If you lose your phone, or uninstall the app, all your saved chats are gone and there's no way to get them back. Android version at least has a useful backup/restore.
Personally I don't think it's likely that signal will close, or that they will sell out. I think the more likely problem is the sort of thing I mentioned, that having a single dev team will be a bottleneck or will reduce user choice. The iOS backup thing I mentioned is one example of that. Usernames rather than phone numbers is another one. Having only one code base does make it easier to audit. And having one foundation in charge does mean there's an easy path to pay for those audits. But it is still a single point of failure.
To be clear- as single point of failure go, I trust Signal more than the next 10 put together. What I don't trust is the whole using phone numbers and SMS verification for sign up. And I would prefer their architecture was a bit more open/federated.