Disclaimer: I'm working on a project in this space. So take what I say with a grain of salt.
There are very few projects doing anything interesting with web3 outside of speculation or "grifter" quasi-MLMs (NFTs). Until that changes, web3 won't be interesting. But, we are on the cusp of that change. Soon, there will be interesting applications on web3, and they will change the web permanently.
Two things stand in the way::
- web3 requires specialized browser plugin or even worse, a special app. You cannot do web3 inside a web2 browser.
- transaction costs and throughput
web3 requires a browser plugin (meaning not on mobile) AND education on how to use it. Those are blockers for 99% of the people, so you only have 1% of the people who will use it, and in that small population, most of them are there to get rich quick or die trying.
And, transaction costs are high for the web3 (I'm equating web3 with Ethereum). It only makes sense to participate for people who are trying to get rich.
BUT: if transaction costs go down, and people can participate in web3 WITHOUT a browser plugin, then people will start building interesting projects OUTSIDE of speculation and MLMs.
That will be the moment really exciting things start to happen.
When will that be?
If you don't know what L2 is, here is why it is important: L2 is a way to scale transaction throughput outside of the mainnet Ethereum. There are many ways to do it, but generally they involve rollups, either optimistic or ZK-proofs.
What's a rollup? A rollup is a way to do lots of smaller transactions off the main Ethereum network, and then roll them all up into a single main Ethereum transaction. You amortize all the small transactions across one single mainnet transaction.
L2 products are still in their infancy, don't work great, and have poor UI. But, they are one of many solutions to solving the transaction cost and throughput problem.
If you are rolling your eyes at ZK-proofs and saying "Oh, those crypto guys, more snake oil!" You should know that Google is using ZK-proofs to guarantee the information they pull from Android devices and processed in ML is has a guarantee of anonymity when shared into Google. There is a lot of interesting research happening despite all the grifting in crypto, so don't miss the good stuff. If you are excited about hard math, crypto has some fun work.
So, L2 will solve the "transaction cost and throughput problem."
The only other barrier is browser plugins, IMHO. The exciting thing is that L2 solutions have their own API. When one of them finallly offers a great UX for interacting with their network, you can interact with it using web2 and not with MetaMask.
Then, everything will open up. It's going to be soon.