Since I train to race, most of my training is structured and I choose a route or the trainer such that I can execute the workout. I upload to Strava to get segment times on the chosen route and I always have power measurement on my bike and upload data to Garmin Connect and Cycling Analytics as well. I do not get Strava measured power and my Strava feed is private.
Sometimes a Strava segment will match where I do a structured element for that days program. That segment becomes an interesting data point for comparison purposes over time.
IMO, most posters on this forum that race, use Strava to hunt KOM’s and that is all about time over distance and not power. Power is ones currency to buy speed. And for racing, speed is what matters coupled with the ability to generate the required speed when it counts.
On highly structured workout days where I do hill climb repeats over / under with rest intervals, my average power and speed will be low especially if I am rolling around during the rest time. Hence, I will log time above 200 watts but not enough to up my average power for the session.
I ride a lot of 4 mile loops around Fiesta Island that is reasonably flat with a good road surface. Stava has no idea whether I am riding on the tops, hoods, drops or riding my time trial bike with race wheels, skin suit and time trial helmet.
With respect to weight, tires make a big difference and the heavier the rider the more important rubber becomes. Stava has no idea what type of tires one is riding or wheels for that matter.
I imagine that under certain wind and bike and rider setup conditions on Fiesta Island, the Strava calculated power number could be way off in either direction. However, lap times in Stava on Fiesta Island are highly accurate and based on GPS. So I find these lap times meaningful as well as lap time average speed. In racing, speed matters.
As a final note on the inaccuracy of Strava on Fiesta Island, I can improve my lap time with the same average power, road and wind conditions by execution riding the shortest distance around the island or riding on the right side of the road for the entire loop (longer distance. Traffic is one way on the island. Also, I can apply more power on the harder sections of the loop (into the wind) which will lower lap times but hold the average power constant. Strava has no way of knowing when I vary power intentionally to maintain speed due to wind.
I think Stava is a great social media platform and KOM hunting is fantastic way to compare oneself to others. Their power measurement algorithm is indicative at best and not very useful for racers doing structured training to see improvement.