Engagement is so interesting to me. The notion that an app can be functionally better but practically worse in a marketplace because it doesn't have enough psychological hooks to keep users engaged. What an interesting software development challenge to face.
On the other hand, once Meet finally ripped off that feature, I basically switched to using it all the time. So maybe Zoom's ease-of-entry was an initial winner, but ultimately there was nothing keeping me there when Google offered the same but without the freemium time limits. Perhaps there is a parallel here with a service like Mumble?
Google's loss was that they had pushed Meet (formerly Hangouts Meet, formerly "Hangouts, no the other Hangouts") as a business only thing while they were pushing Allo/Duo as the consumer apps and chasing the whatsapp/imessage market that year because of their addiction to launching a new chat app every year.
(Hangouts anecdote: In the early days, hangouts.google.com went to the more well known hangouts for users app which had a contact list UI and a different video call interface. There was another link on the same domain to go to the meeting listing/creation screen now at meet.google.com. I don't think there was any way from one to the other in the UI, Google had a g.co/hangouts link to get to it and I think the unshortened URL just had an extra underscore component in the path. I guess they were still on the fence over whether to launch Meet as its own thing or as part of old Hangouts, before they decided to shut down old hangouts)
Later on, it became possible to just go to meet.google.com and click "Start a Meeting".
I've used it and it does indeed work well, but I think it struggles with adoption in part because—it turns out—all those frilly features do in fact drive engagement, help suck users into the ecosystem, etc. Basically it's a repeat of IRC->Slack; people tend to migrate away from things that are easy to migrate away from toward things that are hard to migrate away from.