In my opinion, Voice never had a clear idea for which audience it is meant to be: as you mentioned, KYC doesn't scare off the general (internet) population all that much, but it was a huge no-go for the crypto-natives. For the general population, there was no reason of why to use it. It was not significantly different from Facebook or Twitter, so they had no reason to switch. As far as I know, it also did not have a reward mechanism like Hive, thus people on Hive also did not really have much incentive to go to Voice.

Launching a new Social Network is hard, launching a Web3-based Social Network is even harder, as user onboarding in Web3 is still very difficult. There have been quite a lot of experiments, but most of them failed (just check the last section of Coindesk's report on Voice: https://www.coindesk.com/markets/2019/06/01/blockone-is-launching-a-social-media-platform-on-the-eos-blockchain/)

Just today, I read the following piece on Web3-based Social Media and found it to be very interesting: https://www.decentralised.co/p/volatility-as-a-service. I can also highly recommend this post on the topic, which is a bit lengthy, but absolutely worth reading: https://www.eugenewei.com/blog/2019/2/19/status-as-a-service