There are actually a lot of variables you can look at.
You can check the Kv vs Current, which is great to use for efficiency of a specific motor when setting timing.
But if you want to compare different motors you are also going to need to look at the Kv drop from a no load run to a run with a load. The more the Kv drops, the less torque that motor will have.

With that information we can then start testing and look at the Kv drop under load, and infer from that a torque relationship. Power as said above = torque x rpm. Peak power will be hidden somewhere in the rpm range and this is where you will have to test different throttle levels. We won't have an actual torque value to use, but we should be able to use rpm drop under load as a representation of torque. However, we will have to inverse our results because a smaller rpm drop means more torque. So we will use ((1/rpm drop) x rpm) = power, and remember this is a representation, not actual power numbers.

At this point though, it is all theory. I haven't actually tested the above, I've been pretty busy between my real job, fulfilling orders for my new business, and designing new products. But I do have a new R1 13.5 super short motor arriving today that I plan on testing with. It may take a few days to get to though, but anyone else out there that has the tools and wants to test, I more than welcome your results.