Inrunners are more efficient than outrunners if you compare power and weight, because an inrunner can spin faster, so the gravimetric power density is higher. But to get this higher power density you need to spin the inrunner with really high RPM. This is the reason, why these low RPM/Volt motors have a high voltage, so to benefit from this higher power density you have to use the full RPM capability. We do not only need high power, we also need rather low RPM to be able to use more efficient props, so we need torque. So if you would use the full RPM range of the inrunner you end up with a 2-stage gear and a rather small motor. Unfortenately this is costly and complex. And it also becomes unpractical, your drive system will become longer, which increases the lever and so on.
I agree, that inrunners have a higher power density if used in the optimal way, but to reach this optimum is not easy, especially if you look for efficient large props. On the other hand we have some examples here in the forum, where rather high efficiency seems to be generated with jet-like setups using inrunners without gear. As we want a duct around the prop for safety reasons, we generate drag, as we need a kind of grid at least at the front of the duct to prevent fingers, arms or head to get in touch with the prop, this added drag might not be proportional to the diameter but to the area the prop covers. The larger efficiency of the prop might be nullified by the larger drag of the duct and grid. The optimum prop and duct for a given power output and velocity has to be found first, we dont know it yet.
Another question is, how much power do we need? For foiling it seems to be rather low, for classic surfboards it will remain higher with raising speed.
And in the end its about cost. I can buy a 6384PG 100Kv for around 100-120€ and i have to put some epoxy which creates some work. But than my drive chain is finished. Compared to inrunner + gearbox + sealings + extra bearings + sealed housing + thermal grease +++ this is really cheap and simple. But there are also drawbacks, like you should not use a watercooling for the ESC (e.g. the YEP HV WC makes a connection from Battery to water), or at least you have to make absolutely sure there is no leakage from the motor connections and windings.
The 80A specified is for intense air cooling, water cooling will allow for higher amps, maybe i go up to 90A if needed. The 44V i mentioned in my chart is 12s at 50% battery charge *0.95 because the VESC only puts this out as a maximum dutycycle. I am not sure if other ESC put out higher dutycycle. There are different approaches to explain why, but they all do not convince me so far.
There are a lot of things to think about when choosing between these basic concepts.