I don't usually speak too much about my personal journey, but this hits me square in the center, so I'll give you my two cents worth:
First: NFTs are a way to launder money. Nobody buys a shitty Second Life house model for $800,000 if there isn't an underlying need to show some source of this here cash that I have sitting around. There is currently zero reason for a metaverse to use block chains to prove ownership – it just makes things slower, less reliable, and more expensive.
Second: I've tried, really hard, in my career to build the metaverse, in graphics, physics, networking, art paths, and server technology. I don't think it's going to happen until we have fully-immersive “cable in your brain” interface technology, that's as easy to put on as it is to doom scroll on your cell phone.
I built a part of the technology in There.com. The avatar system there was a lot of fun, and ahead of its time at the time, and the art path in 3ds Max was pretty decent, assuming you could afford 3ds Max. I also did networking, physics, server management, and a few other bits and bobs.
I led the build of Online Live Interactive Virtual Environment (OLIVE, used by the US Army, bought by SAIC) While this re-used art assets and networking from There.com, it actually had a new terrain model, a new renderer, and a number of other interesting developments. The full-world full-geometry arbitrary-resolution terrain database we built (with double precision coordinates, but running on consumer cards!) was quite good.
I led several technology teams at IMVU. The IMVU avatar catalog is very large, and I actually like the “lightweight” approach to metaverse, but the community has struggled to break out of its niche look and feel, the art path has been quite byzantine, and the lack of scripting and physics held it back for game-like experiences.
I led the game engine / networking group at Roblox for a while. They have the game and user-creation system down, and a great team, but I think that the blox-ey part of it will prevent most people over 20 from taking it seriously other than as a way to develop games for younger people. This is the closest we've gotten to a widespread metaverse, but I left because I had become disillusioned with the chase for the metaverse.
At this point, I've retired from game technologies for money, and have co-founded a start-up aimed at making sense of all the data that our machines spits out – something a step function better than Elastic, or Splunk, or New Relic. Selling useful tools to enterprises seems to be a better option than trying to detach humans from the real world :-)