Scales yes.

I once checked over several years run times on regular routes. Simply dividing minutes by lbs then using that number to predict time over a route run at a lighter or heavier weight was remarkably accurate. It puts a spotlight on periods where there was improvement in fitness and when performance improved due to weight loss. Most of my runs were on undulating roads but I can predict times over flat courses just as accurately.

I didn't get anything like the same accuracy of prediction cycling, which isn't surprising, although there was obviously the same general trend of lighter = faster.

As we can so easily and accurately weigh and time ourselves over known courses or distances, I don't see there is a need to move from accurate weight and pace data to a 'dubious estimate of power' which does not take into account head or tail winds or reduced drag from running behind another runner and use that to track, plan and execute training.