Yes, because no one ever tried to produce a watch that does more than tell the time before pebble. [/sarcasm]
I do not fully agree with AC and I find he is definitely going too far. But he also speaks some truth. If it seems too good to be true, it often also is too good to be true. The problem is that for most people, the concept of risk in design is quite abstract. As an engineer, I can weight the risks not only because I know how to do that and have experience doing it, but also because I understand the technologies and problems linked with project management. So when I support a KS project, I have an idea what the risk level is and how good my investment is placed.
Now, most people cannot do this for one reason or another and their decision to invest is solely based on enthusiasms, thrust and first impression. It may be deeply driven by technological ignorance. The person have a high risk of being disappointed because they have implicit expectations of success for a project which may actually have very little chance of succeeding. This is exacerbated by the fact that project closer to the leading edge (or even to the "bleeding edge"), are those who stir the most enthusiasms and interest, even though they are also the projects with the highest risk. This is a dangerous combination when money is in play, as the investors are not fully aware or informed of the risks.
Investing only is safe and low-risk projects as the AC proposes is a solution, but it's not the best promote incubation of new ideas. But maybe better inform KS users of the risks maybe a good idea. Maybe a open risk assessment could be a solution (a bit like an open peer review of KS projects). The potential investors would then be informed of the potential risk associated with backing a project before they do so. Maybe a project with high risk hoping for 500k funding won't get 2 million USD funding anymore, but that maybe for the best as experience showed.