Utsikt is saying that PEM risk limits the distance that can be walked/struggled during the day - essentially a 'step count' limit.
Heart rate monitoring presumably focuses on not taking the rate of activity too high.
This is mostly because cost is relative to budget, and accuracy is more important than precision. Unless it needs to be very precise, but even basic accuracy still needs some level of precision, of counting exactly, as otherwise it's just guesstimating.
Relating to a budget, someone whose income is always below basic necessities will have to count every dollar. Meanwhile someone who has some slack in their budget may only take a cursory look to make sure they don't deviate, while following a rather strict budget, while people in higher brackets can pretty much stop bothering beyond merely glancing every once in a while. Or at all, if they are fully wealthy/healthy. Precision matters less as it grows, because each absolute increase has a lower relative value.
Meanwhile someone who is far below a poverty wage, someone living with severe ME/CFS, not only has to watch for every dollar but every single penny, and even has to precisely take into account future budgets, because for them going over by $2 may mean a 20% relative increase, whereas to someone who has a low but livable budget that $2 would be more like a 2% relative increase, and that 20% offset eats into the next cycle's budget.
The precise numbers don't really matter all that much, but they can't be accurate without being at least precise enough that they form an accurate picture.
This all makes tracking as much art as science, because the real numbers don't mean all that much, and nothing can be put into simple values. Especially as we start improving, all the old relative values get distorted, and this is one way it's so easy to "over-do" it, even if "over-doing" it means the equivalent of eating an extra slice of bread every day, a trivial amount.
Because functioning is not a linear scale, it's exponential. Humans are terrible with exponentials, because we mostly work in relative terms. I've been improving somewhat over the last 2 years, and recently reached a stage where I can start doing things. I can now do more things in a day than I used to be able to in a month, so my calculations for the relative cost of something are wildly distorted. Cents don't matter much anymore.
This makes it very hard to estimate progress in recovery. A year ago I was thinking my progression was maybe in the range of 20-25% of my normal self, but seeing where I am now, still so very far from normal, yet capable of doing so much more, it's obvious that I was a lot closer to the 7-8% range. It's one of many reasons why scales like the CFQ aren't just useless, they're downright foolish.