I have Elite Qubo power fluid and I recently purchased a Power2max powermeter. Yesterday I used the trainer + speed sensor as a power source for the Zwift as usually and in parallel I recorded the Power2max data to my Garmin.
Afterwards I have compared the Zwift and Power2max power curves and I see that the Zwift power is off by 100W after a 45min training.
I also added the “difference” curve and it seems to me, that the longer I ride the bigger the difference is. My best guess is, that as the fluid in the trainer heats up (to about 150 degrees Centigrade), the friction is lower and so the resistance decreases. Howerver Zwift uses constant power curve so it does not take it in account.
For me it meas I did not train according to the plan last couple of months.
Of course I will now switch Zwift to read power from the powermeter, but I wanted you to know that the virtual power with this fluid trainer is unreliable. Maybe you can implement some dynamic power curves for fluid trainers?
Here is the graph. Look how the difference is constantly rising.