I agree and have come to same insight over time. Yet after new development is available, there’s that bit of extra effort that leads to the adoption of the innovative aspects of it. And here FOSS projects usually do not go further than their own project’s scope. It is not seen as worthwhile, win-win, or there’s a “someone else may do that” attitude.
I think not much is needed to bring enough of a change that speeds up fedi evolution, and avoids deterioration into ad-hon ‘non-interoperability’. Big coordinated communities certainly won’t work (imho), and big formal steering bodies may work, but likely not in favor of The Grassroots Fediverse in the long run.