Sorry for the delay.
I can check to add a revision 2 of the power curve with the one you posted. Let me know if you are able to convert the formula to a polynomial of degree 3, usually the formula are using this format.
i.e : power = speed^1 x constant + speed^2 x constant + speed^3 x constant
I think contacting the company would be a good bet to get the exact formula instead of deriving it from a small graph in the doc.
And I'll add. As far as the system knows its is going x mph which = y watts. But changing the tension does change the resistance that you feel. If you had a power meter, it is upstream of the roller/resistance unit and would show that you would require more power to go x mph when more tension is applied. It's been posted on forums that doing a coast down check is the best way to keep the tolerance narrow. I have a KK fluid trainer, I read about and started doing a coast down check many years ago. KK has published their resistance curve and based on guys with power meters they found the following: When the unit is cold spin up to 20 to 20.2 mph and coast while timing with a simple digital watch the number of seconds it takes for the wheel to stop. Cold it would take ~12 seconds to stop spinning, after ~10' of riding the unit is considered warmed up and if you do another coast down check it would take ~2 seconds longer to stop. This ~14 sec coast down time was found to be about the same as the KK published power curve based on power meter readings. Now with the COF2 trainer, not sure what the coast down time should be but if you felt a certain setup feels right then do the coast down and make sure it's in a tight range so that ride A vs B vs C and so on are all comparable to one another. This way the gains are all relative and will show up on the road.
Thanks for sharing! I guess we will need a Power Meter in order to see which curve is closer to the real line. You may want to remove value after 600W to get a better comparison on the graph, thanks :)