The problem is convincing clients that this is the best solution for their problem. I've seen way too many cases where clients wanted (or already had) an overengineered frontend despite the project requirements not warranting it and wasting more time on bullshit such as frontend/backend communication/authentication or build environments than actually implementing functionality.

On the Fortune 500 world that is hardly a problem, Java and .NET backends rule.

Then when doing solo projects, most clients only ask for a site, they have no idea of what they actually want, technically speaking, secret is too deliver something that makes them and their customers happy.

Doing stuff for industries whose main business is not selling software is very eye opening.