To me this implies that if farcaster ever becomes successful it will perhaps happen through alternative clients that cater to niche use cases and thus not as much dependent on twitter-like large network effects. and this might lead to a network effect in aggregate...

I still think this way. People seem to have given up on alternative clients (at least for now and perhaps for good reason). 

That said, I wonder how things could have played out if 10 FC clients competed for various niches (leading to a network effect in aggregate) instead of one client competing with X head-on.