I’ve played with the xert garmin app a fair amount. It does track pretty well with other methods—even when using it ad-hoc. The key with xert seems to be that efforts need to be “to failure” or nearly so. Yes, you can do submax efforts, and xert will give you an estimate. But, if you only do a submax effort for 50% of the duration you “could have” then xert will underestimate your threshold. That said, worst case is that you do a max effort for the max duration you can…its hard to do more than you can.
So, What I like about xert is that it seems to be difficult (or impossible) to OVER-estimate your FTP. At least that’s been my experience with it thus far. I wouldn’t be surprised if there’s a degenerate case or two—possibly involving some short VERY hard efforts. But, for me…it has always been a few watts lower than estimates using other means…assuming a reasonably steady state effort. For an evenly paced the estimate will slowly increase to your actual FTP value. So, if you stop early…or, try to go to hard…you’re more likely to end up with an estimate that is a tick low rather than overshoot.
Thus, you end up with a conservative estimate of your FTP and training zones—which is a good thing.
The app makes a rough estimate of the anaerobic contribution to your above threshold efforts. Like the 95% rule for FTP tests only scaling it depending on previous data (we do know that the 95% rule doesn’t apply to everyone, right?). So if the app is underestimating, and the efforts are exhaustive, it may be that whatever you’re using to estimate it is overestimating instead, especially if there isn’t a quantification of your anaerobic capacity (High Intensity Energy or HIE in Xert terms).