I think it would makes much more sense, to have a dedicated project which is focusing on just collecting the data and have a community to maintain updates. Then all those frontends could connect to whatever API this would offer.

There are many fancy projects like this for high profile environments in companies or specialized areas (like home assistant), but they all are very cumbersome to start and maintain. Something simple which could run with a oneliner and some config-file, in good old unix-mindset is really missing in this space.

> think it would makes much more sense, to have a dedicated project which is focusing on just collecting the data and have a community to maintain updates. Then all those frontends could connect to whatever API this would offer.

It already exists: https://woob.tech/

Yes, something like that, but not like this. This is too inflexible and limited. The datascrappers here are written in code, which makes it impossible to exchange them with other projects. And it seems there is not even a server which makes requesting data from an external process possible, to feed it to your own code. There are many tools like this, with a collection of specialized datascrappers for their own demand, and they are all on their own, isolated from the rest and everyone is constantly recreating the same.

It is code, so it is the most flexible that can exist. You can call the code with an application that includes it, you can make your own server with your own API although that isn't really useful if the server is going to be local anyway (most of these thing deal with private data, you don't want a central entity managing private data). You can call a process from another process, I don't see what's stopping you.

Perhaps if you have an example of what you want to build I can understand how woob doesn't fit what you want to do