What factors, other than cost, might be hindering Firebase's adoption as core infrastructure for the web?

Well, the really nice APIs that their SDKs have are nice, but ultimately, fundamentally what's below that is a NoSQL document database, so you're limited in terms of the types of queries you're able to execute. Whether the API is nice to work with or not, it's not going to be able to change fundamentally what you're working with.

These types of data stores are designed for extracting very specific workloads that have different scaling requirements of properties than the rest of your app. You can put something in a DynamoDB and just scale it to infinity independently as one piece that you've extracted out of your app, but to build your whole app with this kind of a database is really limiting, or it can be.

That's where I'd throw in a plug for Supabase, which is kind of taking the developer community by storm the last year or so. Supabase markets itself as the open source alternative to Firebase, but it's actually fundamentally just a different technology. It's Postgres over the hood; it's a relational database. Then they're trying to push that ecosystem forward with a more Firebase-like developer experience, with really nice SDKs to work with, really nice APIs, really nice tooling and dashboards and all that stuff. It’s really easy to get set up and get started. So you're seeing some of these things come together to have a real set of options to build real, scalable and sophisticated apps with a Jamstack model.