Thanks for the additional info about your personal experience w/PMs. I tend to get pretty well informed before I make important decisions and I consider spending $1K+ an important decision. Since I did my serious cycling before the availability of HR, cadence or PMs, and was reasonably successful, I don't hold out any bike computer info as the "holy grail." But, since the objective of cycling is to apply as much force to the cranks as possible for the required duration, it is pretty obvious that it would be useful to know how much power one is applying to the cranks at any point in time and averages over time. I understand the technologies underlying all four of the available products (I have read the patent applications), and I think I have a pretty good feel for the technical and maintenance shortcomings of each. I see a lot of discussion about accuracy, but to the extent that we're talking about 1.5% or 3%, I can't figure out why I care, so long as my device is consistent. If 200 watts is always 200 watts, it's fine with me, even if it's really 205 watts. Intermittent readings are, however, a problem.
As to training with a PM, I don't know that it will change my approach to training very much except that I will probably define my rides in power terms with a spot check of HR. I have a pretty simple philosophy for training. I ride for about 3 months mainly spinning whatever gear I can turn without feeling like I'm pushing hard. Then I start doing hard days followed by easy days. Hard is hard enough to make my thighs and lungs burn for at least 5 minutes at a time, sometimes more. Easy is easy enough to have a conversation with a riding buddy. The main difference is that with a PM I'll know what hard and easy are in watts. Whether that'll cause me to change the way I train, I don't know yet. But, there's no way I'm going to have a training schedule that looks like a NASA launch schedule.