Disclosure:
I feel strongly about this one and figured this might be a fun place to put some more general thoughts on the interwebs. I mean no specific harm to the author of this thread, or of that paper, or any companies specifically involved.
I wrote two versions. Neither were edited or revised as thoughtfully as I have considered this subject. Pardon any lack of coherence, spelling, or grammatical issues.
**TLDR version: **
Will it work for some folks? Sure probably. Most things work for some folks. Is it the next step in the evolution of scientific training? Probably not. It could be just another one-dimensional approach to add to the growing list of tracking-first, thoughtfulness second, training approaches. And there is possible harm from growing that list.
Rant version:
This novel approach is another “tracking first, thoughtful intervention second” thought paradigm. We need to stop letting training paradigm hypotheses, and all training and nutrition thinking, be dictated primarily by what we now have the capacity to measure. If we do not stop these lines of thinking, and think more critically about the whole, we will be increasingly misled by the now-exploding ability to measure new things.
The ability to measure something does not confer any special value to the measured thing, no matter how accurately, continuously, or granularly we are able to measure that thing. Nor does the ability to prescribe interventions using that thing as a measuring stick.
The fact that some of the earliest “things we could measure” were indeed quite meaningful in terms of training & nutrition prescription (aka ‘intervention’) misled many would-be inventors over the coming decades. The reason the early measured things changed the game in training was because they were closer to fundamental realities (basic physics), less reductionist (looking at overall body systems instead of singular molecules and concentrations), and because someone would only invent something to help measure something if they were pretty confident that that thing was going to truly enhance performance. Now, compared to 20 years ago, the cost of invention is lower, the capital has been flowing loosely to get things in front of broad markets quickly and publicly, and we’re seeing an explosion of “lets measure an alphabet soup, throw it at the wall, and see what sticks” approach.
The result is myopic recommendations based on overly reductionist tracking and measuring of often misleading biomarkers and other ‘variables’ which can be measured, graphed, derived, or otherwise correlated to various and sundry other things.
The “graph” is a hot-button seller. Everyone wants graphs. We want to see ourselves visualized. Analyzed. Measured and assessed. But before purchasing something to see a graph, we must ask ourselves: "if that graph were to tell me exactly and reliably X, and it told me to do exactly and clearly Y, precisely because of X, do I have any evidence that doing Y would actually make me better off, in light of the things that doing Y might otherwise replace or cause me to lose? Here are things you might lose because you decide to do Y: freedom, time, other training characteristics or emphases, training variability, enjoyment of training, money. There are more. Each of these often independently outweighs any potential value of doing Y, when Y is based purely on a reductionist X.
At least we must not delude ourselves that just because we measure that thing very well, we will therefore be better off than if we did not measure that thing. Among the nerdier crowd, including me in my younger years, there is often a pronounced placebo effect simply from measuring more things. This is not true for about half the population. I’m not opposed to self-measurement for fun and the personal enjoyment it might bring to some of the nerdier among us, or for leveraging the placebo effect to your advantage. Just take care not to delude yourself in the process.
I *am *deeply opposed to measuring things automatically making people better. It does not, and it very often, especially among people who will never post on this beloved slowtwitch forum, makes people worse. Even the mere knowledge that other people are able to measure, and are measuring, more than someone is measuring, is enough to have a negative placebo effect on folks who do not love analyzing data in their free time. Sometimes this negative effect - of either tracking, or the awareness of one’s own decision not to track oneself more comprehensively - can be quite strong and it is certainly not limited to a small subset of the population. It probably affects more than half the population negatively. Their awareness of such negative effects may not match the reality of the negative effects.
The take-home: if you love data and love analyzing data, nerd out. Track and measure to your heart’s content. Just don’t delude yourself in doing so. If you do not love data and analyzing it, take heart, you don’t need to start using fancy tech to be tracking lactate, or blood sugar, or sweat rate, or heart rate variability, or sleep stages, or anything else. And you shouldn’t. You’re genuinely better off if you don’t, unless a medical provider who knows you well, and is looking at the whole you, says that you definitely should start measuring something regularly.
/rant
Mostly saved, in case you edit in the future.
The more you post…the more I like you. 
“Why are you looking for your keys here? Because, this is where the light is!”