Surely, you've heard of Octopart, 10gen, Blekko, DDG, Heroku, Hipmunk, DirectedEdge, RethinkDB. These are just some off the top of my head. If you don't think they're solving hard problems (not in the sense of sharing lolcats at webscale), then we'll just have to agree to disagree.
Oh, no, some of those are the companies that I am the most excited about in the current tech scene. I am not the kind of guy anyway who want to prove a point on this (despite the above 'rant'). I am just happy do discuss this. (I am an INTP and not INTJ). I don't have data on this, I just have a feeling that there were way fewer so called tech-startups 30 years ago, and according to some metrics (revenue, publicity) the ratio of highly-tech-oriented-from-the-start startups were higher. A thorough research on this would be interesting.
But I don't care about them now. I don't care about customer aquisition. I don't care about reelase early-release often. I am an engineer, and I am better at it than any of the self-help gurus out there. As an engineer I have my own internal thought processes which I have refined for years. I research and develop ambitious and deep technology. I think about the technology under the shower and on the train. Graphs, hierarchies, O(logn)-s, objects, classes and functions, (even conceptual) refactorings, sometimes even axioms and theorems are dancing in my mind all day. I have lots of TODO-s. I have new ideas every day. I know what to improve, I just need time, and I will improve those things. My motto is: TODO: 'optimize this further', 'TODO: figure this out', 'TODO: think about this from a new perspective', 'TODO: consider refactoring this in xyz way'. (I create a new GUI system in Javascript completely drawn on HMTL5 canvas. Kind of reinventing RIAs.) Engineering is important and brutally overlooked nowadays. Suprisingly few startups do ambitious deeply technical things.