When it first came out I logged a highly structured test ride on both a powertap and powercal at the same time, just to see how much it might be off.
And also 75minutes of 'riding around', no structure.
Interesting data. But there are a lot of relatively high-power/short-duration intervals in both those examples. Exactly the types of rides the PowerCal won't do well on.
I'd like to see 2x20 where your NP is not much higher as your AP - i.e., no 1 kW 5-sec power excursions that drive up NP and therefore IF and TSS but don't really do much otherwise.
Then compare the two outputs.
Heck, how about a 90-min z2 ride on a trainer? AP = NP = 180W, w/ max 1 sec power 190W.
Why do so many seem to have a need to throw rocks at a relatively inexpensive training aid that looks like it can provide 90% of what a power meter does for less than 10% of the cost? I own an SRM and a PT, and I don't feel bad about someone getting a tool that for $100 provides 90% of the data I get from devices I paid a helluva lot more for. Good for them. Knowing that my max 5-sec power is 1724W does what, anyway? Tell me I have a good sprint? Yay for me. No one needs a power meter to figure that out.
If a PowerCal provides consistent enough data to allow accurate training with power over intervals, say, 5 minutes long or longer, it provides data good enough for it to be at least 90% as useful as a PowerTap, Quarq, or SRM.
Because for road cycling, it's aerobic power that matters. FTP too low? You're dropped. FTP high enough? You're doing the dropping.