First, the two parts of your advice aren't identical. That is, "something people want" doesn't always mean "something that solves a problem."
Second, you're being overly demanding, I think, even about "solves a problem." We could do a whole song and dance about the human need for fun, but that would be beside the point. When you say "It's fun, but I can live without it," you changed the terms of the whole debate. We started with "solves a problem," but tons of things solve problems and yet you can live without them.
Finally, I'm pretty sure that there is a real category "invented needs" (or "invented desires", if you prefer), and that lots of webapps fall into that category. The existence of the product creates the need (desire) after the fact.
Sure I can live without most things that solve a problem, but that's not the point. I want to understand if there's some added value for the users, why would they start (and continue) using it in the first place. Jerome_etienne gave a good answer, but you also made a good point with the "invented needs". On the other hand, I'm not sure that can last for more than a couple of years on its own. Users flock to the next hot thing all the time.
What you say here makes sense, but I'm not the one who brought up "living without" things; you did, in your original question. I agree that the real focus should be on why people flock to these things, but in that case, you answered your own question: it's fun (you said). The value is the fun.
As for not lasting or people flocking to the next hot thing, that sounds about right to me for many faddish things that you find and play with on the internet, no?
Second, you're being overly demanding, I think, even about "solves a problem." We could do a whole song and dance about the human need for fun, but that would be beside the point. When you say "It's fun, but I can live without it," you changed the terms of the whole debate. We started with "solves a problem," but tons of things solve problems and yet you can live without them.
Finally, I'm pretty sure that there is a real category "invented needs" (or "invented desires", if you prefer), and that lots of webapps fall into that category. The existence of the product creates the need (desire) after the fact.