Just to be clear, even if you had more granularity with the power curve, I’m not sure how useful it is. For example, when I ride outdoors, I rarely go really all-out on shorter bits, after all, I still want to enjoy the rest of the ride and not limp home. Indoors, I only do workouts, and they are not designed to push me to the limit. So most of my data is not useful for the power curve (at least not for its intended purpose.
And even if I had a good amount of proper all-out efforts, I am still not sure most people can easily draw useful information out of it.
Maybe we are talking about semantics here, but I don’t think marketing pushes FTP as much as you say they do, IMHO it is that FTP has for the longest time been the only performance metric that is easy to grasp and front-and-center in TR.
(Arguably, PLs still cannot be used to track performance since AFAIK TR does not allow athletes to plot their PLs over time, you’d have to track them manually.)
Why is that absurd? I think it is more difficult to finish PL9 sweet spot workouts than PL5 workouts, and that this reflects a difference in fitness.
The reason why I am saying PLs are not a performance metric — or perhaps a non-ideal performance metric — is that allow no comparison between very different levels of FTPs as well as a number of other, more subtle reasons, most of which are specific to particular energy levels. (E. g. endurance PLs are limited by my time on the trainer, not my fitness.)
Even if PLs were based on full ride data (i. e. including outdoor rides) and it’d accurately reflect your life bar/number of matches at a given power, it still isn’t clear whether a lower FTP and very high PL is better or worse than a 5 % higher FTP and lower PL. Probably that is because the answer is a clear “It depends!”
As far as I understood the official TR messaging, PLs (and workout levels) were meant to fulfill both roles, that of a tool to help you select a suitable workout and as a performance metric. But it was not presented as a performance metric alone.
IMHO TR’s next big frontier, after releasing the scoring algorithm for unstructured rides, are performance metrics. Depending on the plan you choose, they should choose suitable metrics. Ideally, users could tweak the emphasis, too, but that might be too hard for an initial release. One metric that they should implement ASAP, though, is training consistency. Training consistency is probably one of the few metrics that is relevant no matter what your goals are, no matter whether you follow a polarized or a sweet spot-centric plan, etc. It is also easy to quantify and the numbers are easy to understand.