All these running power meters use accelerometers that are calibrated to correlate the deceleration upon foot strike to running pace. That can be fairly accurate when each foot strike is the same such as when running on the the flats. Running on rough trails, uphill and down hill pretty much means the algorithms used to give power will not work. The new Polar power meter claims to correct for ups and downs by including a barometric altimeter.
My experience with the Styd (3 years ago) left me unimpressed for mountain athlete use. When going uphill and working hard (but with low impact force, it had lower power readings then an easy running pace on the flats.

From my, admittedly limited experience I would say that HR is still a better proxy for training load for our variable training conditions.

Scott

They have worked fairly dilgently on updating the Stryd algorithms to account for “power hiking” or walking steep inclines. I’ve been using a Stryd for over 2 years and find it now does a pretty good job of correlating effort when inclines necessitate a walk in order to stay in zone.

On relatively smooth (i.e. non technical) trails it can almost always predict an HR spike before it happens. This is especially evident post-run in reviewing the graphs.

Where it still has difficulty on climbs is when the trail is so technical that it resembles stairs or step ups, it completely drops the ball in these situations.

I find that it’s still a useful tool however I have to take into account HR and perceived-effort in addition to the watts it’s measuring/displaying, especially when racing.