How are you measuring these speakers?
- Mic 0.5 - 1 m distant?
- Speaker away from close reflections in the room?
I ask because the direct sound is best measured this way.
It appears there is lots of room effects in your measurement. We don't want that if we are looking to identify the phase of the direct sound. You could apply a FDW of maybe 4 - 6 cycles to your measurement and see if the phase is improved enough to see the trend of the direct sound phase rotation. A new measure is probably a better idea.
Personally, I use an FIR filter to correct the actual measured phase rotation rather than the excess phase portion. It does seem however, that it is much more popular among other hobbyists to correct just the excess phase portion.
Below is a 32" measurement prior to any phase correction. the DIY speaker measured is shown in my avatar. I do have some nearby room interference here, but selected REW settings to better reflect the direct sound.