Ooo, nice. This is quite clever - still will require insight and effort from the rider, but it should be less of a shot in the dark, if it works. I'm not in your target market personally, but I like the approach.

Are you going all out with triaxial load in the seat post, or uniaxial (posterior)?

You're asking the right sort of question :)

Seatpost is actually one of the simpler sensors, without giving too much away it's sensing force in 4 different directions.

Is the 4th direction time?

Typically it would be a rotation (or torque, if we're talking load force/torque cells).

Limited in what I can say, but you're not wrong :)

I don't understand why you need to measure anything but instantaneous crank torque, speed, and (for completeness) pitch angle. Integrating power in vs acceleration out, the difference is loss. What am I missing? Does bad posture waste energy before it reaches the crank?

In an ideal world with zero friction, zero wind, constant air density and no other losses that would be true. There are a few outfits using this approach, so we are the outliers (being the only ones doing it via direct force measurement, for which we have the patent).

The difference between energy in and work out is, exactly, friction, wind, and "other losses", so you have not said anything meaningful. That is your right, but then why answer at all?

Depends if you want to answer “what” or “why”.