We already know that, and is not addressing the question asked.

It's pretty simple. Either the P2M is:
i. measuring the crank rotational velocity one per revolution and using that to calculate power, or
ii. it's measuring crank rotational velocity many times per revolution and using these more frequent crank velocity samples to calculate power.

If it's the latter (ii.), we'd appreciate more information on that, at least what you can say without release of proprietary information, e.g. how many times per second or per revolution is crank rotational speed sampled, and to what level of accuracy. Although we might be interested, we are not asking how P2M might do that, which would be the proprietary information I'd have thought.

If however it's (i.), then we know that the P2M will have the same accuracy issues with non-round rings as other crank based power meters, and a claim that accuracy is not affected would be incorrect.

Further, if (i.) applies, I think it's important to then explain the accuracy of determining crank rotational velocity once per revolution, as if it checks say only every 1/50th of a second, then to help understand how the P2M avoids a sample by sample precession "error" in power calculations.