And you'll notice that the great many of these cyclists are continuing to train effectively nonetheless. Don't get me wrong, I'm all for nerding out with data, but the reality is that the big picture is that it doesn't really matter unless you're using some very advanced techniques. The fundamentals of training don't change with 1%, 2% or even 5% data errors.
I've been using a powermeter since around 2003-started with a square taper srm. Have had a number of others, and am currently on Stages. I don't doubt that if I was conducting a scientific study of my physiology it might be less than optimal, but I'm not, I'm simply training to be an effective cyclist. For that purpose, it is perfectly adequate. The price point has allowed me to have a powermeter on a couple bikes, whereas with a crank based one that would be stretching the budget.
So long as you consistently use the same device, and use that device to test, AND the device is consistent, then you'll be fine. I'm aware of the argument that a 5% tolerance could conceivably lead to a 10% swing (5% in both directions) leading to say, a 20-30w inaccuracy, but in my experience that's not how these thing work. The errors tend to be somewhat repeatable. Weakness are identifiable-stages are known to have somewhat of a laggy response to very short efforts, so maybe less than ideal for track sprinters. I am well aware that Doc Coggan and others would disagree, but the level at which they're used to analyzing data goes well beyond where I'm at. Anyway, its more or less the longer term trends that matter in power training.
And, we've beaten this to death. The study cited isn't really new information. Folks on either side of the debate aren't going to convince each other. But I'm not regularly getting trounced by guys on SRM's because they're on SRM's.
Completely agree with you !
Pro need precise powermeters. If you're not a pro, most of the time the goal of the powermeter will be to find the good speed during a climb, because a regular speed is the warranty of the best climbing time. If you start at 600W and finish at 250W, it's not a good deal, a regular 400 is better anyway. That info will be given by most of the powermeters of the market.
When you don't know a climb, you just put the numbers of the Strava KOM and make an estimation of your speed depending on your power curve.
For me, powermeter is a speed regulator and you can be stronger than other competitors, just because you know exactly where the red zone is.