There's a cause and effect at play between the blades taking power in and the generator sending power out as electricity.
These two powers need to be in balance, otherwise the blades speed up or slow down, seeking equilibrium. The seeking of equilibrium isn't particularly concerned with the arbitrary choice I made to carve the blades for any particular TSR. It will happen even if, to maintain equilibrium, the blades must speed up while moving away from the TSR intended in their carving. Larger blades always provide more power input, as long as other characteristics like chord, stiffness, and airfoil aren't also changed for the worse. Putting larger blades on the same generator usually makes everything turn faster. I've done the opposite: putting a much bigger generator on the same set of blades, and that's why it runs too slow.
The goal with carving a particular TSR in a set of blades is to hit a spot where they actually ARE at the design TSR when this equilibrium is found between energy at the blades and energy to the generator. That's what gets you a high peak CP. What you're seeing on the chart above is what happens when the blade speed is not matched to its design TSR. If you went back to previous posts I made last year, you'd find that the TSR did match the design, and the CP was very high. That's what swapping the generator has done.
The simplest choices in front of me are (a) carve a bigger set of blades, and choose a TSR curve to match the diameter with the generator's power curve, or (b) install a MPPT controller that will run the generator faster, allowing the blades to function at their design TSR for peak efficiency. Since things are working fine, and safe, for now, I have no need to change anything right away. Nothing to complain about because it cuts in OK and the peaks are spectacular. I can collect more data and plan which of the two will be the most effective (or most fun).