Running power meters was covered on the DC Rainmaker/ GPLlama podcast this week, as Ray was testing the new Stryd which will take into account wind (which the earlier one’s didn’t). There was a bit of a throwaway discussion about how there isn’t actually an agreed running power calculation. So the stryd is probably consistent, but there’s no real measure of accuracy (was my take anyway).
fwiw, in the calculators for TDEE, I use the somewhat active, as I generally get the 10,000 steps regardless of runs. Garmin Connect gives average stride length (for me, using HRM-RUN anyway), so I guess you could take away running steps and see where that leaves the step counts?