Does this make ETH useful in the real world or is it still a tehcno-ponzi scheme?
But on paper building something useful now is a bit easier as transaction cost is likely to come down and the network should support a more reasonable transaction volume.
When your global transaction volume per hour is capped below what even a small web server running on a tiny computer would easily handle, there isn't much you can do with it in the real world. And when those transactions then cost you tens/hundreds of dollars and take minutes/hours to clear, any utility that might exist goes out of the window. It's a complete non starter for anyone looking to do some transactions that actually have business value. Why would you? It's many orders of magnitude worse than what a plain old database gives you on all relevant dimensions.
That was the status quo up until the merge. Now that that is in the past, we'll see. I personally doubt Ethereum will be the tool of choice for this kind of thing. If you are serious about building something that does something useful, there are better tools available.
ETH was already potentially useful in the real world if you were on an L2 like zksync - low fees, fast enough to work, decent UX with wallets like Argent. Only thing missing was wide adoption.
Having said that, the merge is actually just the first of a multipart set of upgrades that should make the L2s massively better than they are today. We're still at the 'potentially' useful rather than the actually useful, but things improve the whole time.