This is sort of on my list of things to do. One side benefit of (at least the Raymarine) modern autopilot is the flux gate compass sensor has been replaced by a 9 axis MEMs sensor, and all of the effort needed to develop artificial horizon software was long ago sunk by the drone and cell phone folks. So roll, yaw, pitch, and heading are available at a very high frequency on the NMEA 2K bus. As well rudder angle. There are now cheap devices that record this continuously. My boat is clearly not an UL lifting keel, but it should be easy for someone to collect the data. Go out in the next F8 with your Pogo, run for awhile keel down, then keel up. You could compare roll and raw excursions against rudder angles and probably learn something.
For the ultimate question, we would need 100 volunteers in 100 Pogos, half with keel up and half down, running in extreme conditions and see how many in each reference class die. But that is an experiment unlikely to get funding.
I'm a little curious about the structural integrity of keel up operation. Not familiar with the details of the keel mechanism, but for some that I am familiar with, keel up would not be great. For that matter on some keel down isn't good either, but that is a separate keel integrity discussion.
For the ultimate question, we would need 100 volunteers in 100 Pogos, half with keel up and half down, running in extreme conditions and see how many in each reference class die. But that is an experiment unlikely to get funding.
I'm a little curious about the structural integrity of keel up operation. Not familiar with the details of the keel mechanism, but for some that I am familiar with, keel up would not be great. For that matter on some keel down isn't good either, but that is a separate keel integrity discussion.