Here's the opinion of someone who thinks more about the ecosystem of solutions than the individual solutions, that would be me.
One thing I think is so often a missed opportunity in early stage FOSS projects is collaboration on code base. In my mind the best solution must be generic libraries (preferably in a popular high level language) that are easy to integrate into applications. This focuses efforts and skills and makes it easier for end users to find a small number of implementations in (preferably) popular applications. Now, I'm not knowledgeable enough to comment on any specific project and the pros & cons, but .dxf & .dwg support are classic examples of a fractured solution with a lot of incomplete implementations. Likewise the proliferation of implementations of libraries for handling IFC data. Of course sometimes a different programming language may be a prerequisite for implementing some specific solution.
For me personally I live in a co-op and support the political cause of employee & customer ownership of businesses & infrastructure. So this is not some passing thought I had riding the bus home. OSArch for me is an extension of this principle. Power to the users & cooperation whenever it makes sense.
So how much of this makes sense to you? Am I missing an important perspective?