Overall, I’m pleased with how our series turned out (but I’m definitely glad it’s over). It was also nice to have the more practical or ‘on-the-ground’ perspective from TrainingPeaks.
On that last note, I’m thinking that the following comment from your last article (excellent series, BTW!) is a bit short-sighted by the coaches you talked to:
After discussions with several professional coaches and physiologists, all agreed that a person can adequately train to the data from any of these systems. They all noted that the most important thing is that a power meter is accurate relative to itself.
What happens if the power meter breaks? Now you’ve just called into question ALL of your previous training data in relation to your current data. Are those personal power bests you just set TRULY “bests” and a function of your recent training, or are you just getting “fooled” by a new power meter that’s “accurate relative to itself” as opposed to being accurate overall? You’re suddenly not able to maintain power levels you could before. Are you overtrained…or is it just the new meter?
As I’ve said before, if a power meter can be “accurate relative to itself”, there’s no good reason it can’t be accurate in an absolute sense. That’s called “calibration” (and NOT in the ANT+ misuse of the term to mean “zeroing”). It’s really not that hard. I’ve done this with many Quarqs besides my own using the Qalvin app, and nearly universally the response from the owner ends up being “Is that all it took? That was easy.” 
IMHO, if Stages was smart, they’d implement the ability to check and modify the torque slope of their device. I have to believe it’s possible with their unit. For folks with a relatively consistent L/R balance with effort, this would allow them to “tweak” it into accuracy with a “known good” PM that measures the output of both legs. I did this at one time with an Ergomo (i.e. modified the slope) and got it to read VERY close to a PT.
“is a bit short-sighted by the coaches you talked to:”
I agree. Historical data is a big deal. In general, I tend to hear more of the “accurate-relative-to-itself” sentiment from manufacturers and physiologists, and “absolute accuracy” from athletes. The athletes seem to want to know the history and trending - and don’t look at it just as a training tool. It is a diary-keeping tool, too.
Beyond the accuracy of the PM, I think this also really shows how important the head unit is, and the software that you’re using. I’ve seen as much variability due to the latter two as the PM itself.