I just read your "fallacies and failures" article and wanted to say that I thought it was terrific. I am now a veteran of four startups in four different industries, and at the risk of admitting that I must not be a quick study, I think we made similar mistakes across the board. So I think that the tendency to forget about some of the fundamentals has been around for a long time, it is really not a brand-new phenomenon by any means. Remember pen-based computing, or all the PC companies that started up in the early '80s? I'm dating myself ;-)
The Internet has just made it more "acceptable" lately to shoot from the hip -- after all, it ended up working for quite a few people, even now, when a lot of investors are taking a bath on ill-conceived projects doubly cursed by poor execution. Take the Value America example, chronicled in Business Week. Mr. Winn is OK, he took 60 million or more out, yet a lot of folks have lost a bunch of money in that fiasco. And even a cursory analysis of what they were trying to do would have pre-empted a lot of the problems earlier.
Why were there a bunch of petfood companies funded at the same time? No good reason. Why did Drugstore.com and PlanetRx not understand the whole issue around pharmacy benefits management until they were well underway with their businesses? No good reason. As you pointed out, why would people conclude that e-books would be adopted like cell phones? No good reason, but that's a more complicated and perhaps mildly sinister story associated with the unholy "professional research" topic!
I could go on and on, but I've probably gone on too much already! Don't even get me started on the whole "internet appliance" debacle to come!
To reiterate, great article. The problem is not limited to startups, as your Informix example showed. In fact, large companies frequently have enough resources to let them fail slowly. They also have existing customers to call on, many of whom try to be polite when presented with the latest "cool" thing, infoblades or whatever. So the sellers walk away from the meeting trying to tweak their offering instead of realizing that they are selling stuff that people just don't want.
A lot of the problem is related to human nature; people want to be optimistic about their endeavors, and don't want to hear bad news. The trick, I think, is how you can drive an organization towards rationally examining what it is doing without being branded a "dark cloud" or worse. I wish I could say that I have it all figured out, but I don't by any means. I know a quite a few things that don't work, unfortunately!