Not to downplay any of your points, but as a counterpoint, I'm not sure this was true for Vim/Neovim. Perhaps I'm only looking at short-term gains though and not long-term fragmentation
On the contrary. Emergence on neovim had a positive impact on vim speed of development, and Bram's attitude to new features (vim terminal, async, etc) softened a lot. He has also become more open to contributions from others. Even though vim-9 and neovim are on divergent paths they are both better because of the competition.