The fan I added to the battery box made a world of difference. I was able to run at full power for close to an hour, with air temperatures inside the case maxing out in the mid-90s (mid-70s ambient). Without the fan I would have been over 110F. And the positive terminal is now cool enough to touch, compared to before where it would be too hot to touch after only 15 minutes of full power.
On the hulls, there are more efficient shapes for sure (Wintech, Stillwater, ...). One of the nice features of the RS hulls is their toughness. Hitting a rock at speed is not as likely to damage them, compared to a fiberglass hull. I sacrificed some efficiency for toughness:
Note that these hulls are considered as 'semi-displacement'. So 'hull speed' limits are less applicable. From Rowing Solutions I was told that speed is related to power by square root (p = Cv^2). This matches my observations well.
Difficulties in obtaining a power-to-speed table include variability of weather/water conditions, and different loading levels. It turns out that this style of hull is more sensitive to loading levels, compared to a monohull style.
What I can report, from my most recent outing (Shilshole to Edmonds, 22-mile round trip), is that 2/3 power easily yielded 10mph, and that's with a fairly full loading level (backup transom/motor/seat installed, for example). By solving for C with those numbers substituted, I get this 'theoretical' table (C = 40):
1kW = 5mph
2kW = 7mph
3kW = 8.7mph
4kW = 10mph
5kW = 11.2mph
6kW = 12.2mph
I've been in the 20.x km/h range on a good day (smooth water, light loading), so that's close to 13mph. And I've seen 19.9 km/h with more typical loading/water conditions (right at 12.5mph). So the above table seems close to reality, at least regarding the 4kW and 6kW lines.