1. A low-volume competitor will always seem better because the bad actors haven't moved in yet. Growth engineering will kick in and do the same awful thing you see on any newcomer that approaches or surpasses whatever it aims to replace. Put another way: the quality will only ever go down; and
2. We're going to stay in this cycle as long as we back venture-backed companies that will be incentivized to do all the growth engineering mentioned above.
The most successful and enduring and longest-lasting structure for user-generated content is Wikipedia ie the Wikimedia Foundation. The platform needs to be owned to the community or we are doomed to repeat the rise-and-fall cycles we have now.