Power would be force and time. I’d argue there would be some impact, but not a huge impact. On average, it would be fine. There would be increases and decreases and some intermittent spikes. But, the same can be said when I ride over rough pavement.
Also, there may be some power abnormalities, but most any software can filter out abnormal spikes below a certain threshold of time, and ignore it. I know that Stages L only would get spikes off-road. Someone I know goes through that all the time, when they ride L foot forward on descents and get power spikes when there was no cadence.
I haven’t seen that one. I’ll look for it, because I’m not sure what it’s referring to. Thanks for the tip.
They haven’t advertised that it’s “smooth pavement only” as well. Potholes? Rough roads are also invalid and “shall not be used”? I looked at their site and the user manual, and it’s only about pedals with a particular cleat. There’s no mention of smooth pavement only or any of this.
At the end of the day, it’s just a tool for measuring power. It should work wherever, in a reasonable manner. Reasonable being the keyword. If the argument is that the forces on the pedal can vary over rough surfaces, then so can forces on a chain, and therefore the same argument applies to all power meters. As someone who owns multiple meters, they all seem reasonably capable, on or off road.
So, I don’t buy the argument that you shall not use it on gravel. I think it’s more about the secondary implication that it probably involves the SPD hack.
Edit: Just to make it clear. I know people who use road pedals on gravel. There are plenty of people in my cycling clubs.