Not all the investments of the dotcom bubble were malinvestments. A part of it has been re-used, but another big part was wasted. There was no need for all the crazyness and all the waste to create the infrastructure we have now.
But this applies to everything. Even in non-boom times, a large part of investments is wasted. I've come to see it as a kind of decentralized, randomized discovery process. A boom/bubble is just a speedup of the normal rate at which this happens.
On the other side you have the structured, committee-based economy, which tries to produce growth at a fixed rate by guessing what things are real and useful and which are bubbles and shouldn't have resources wasted on them. The problem is, that it is hard or impossible to know those things in advance. Good ideas might be shot because "they're just a bubble" and bad ideas might be preserved because they've worked in the past and the world has changed since.
True, the discovery process has a lot of failures, trial and error. All in different sectors, all different things that are being tried, from which only a few succeed. But its not similar to a bubble. A bubble is not in any way a magnification of that process. Hayek always said that what make him some interested in the boom bust cycle was this question: There are normal errors always in the economy, people trying things, some times succeding, but a lot of the times failing (as you pointed out), but why suddenly all the investors and entrepreneurs make all the same mistakes in the same sector?
Its not the same as mistakes in different areas and different ideas. During a bubble the investors and entrepreneurs make all this mistakes in the same direction. Why? Because they are getting false information (mainly interest rates), the price signals are perverted. And as long as this false signals are mantained, the problem goes on, and that is why there is no correction until even the government can not keep the bubble going with more coercion.
A bubble is not trial and error, its not a discovery process. Its reactions to bad signals, and that is why the trial and error is not called malinvestment (you are getting information), but a lot of the results of the bubbles are indeed malinvestments, and could have been avoided if the price signals would not have been distorted.