Brian,
I see the suggestion differently. If I understand your response, you seem to say that the final outcome of a guided exposure is better evaluated by another controlling app such as NINA. And, again if I am understanding your comments, that the user could monitor the RMS values (RA, DEC, & Total) since they are reported by PHD2, but that there is little use in doing that in order to make PHD2 guiding adjustments ("tweak things").
First, the final image accessible to NINA or other program is the result of however long the exposure is. It is not a moment to moment monitor of PHD2's guiding actions. Second, for those who prefer not to have to watch the PHD2 user interface (UI) real-time, if PHD2's RMS values were exposed, they could be used by the user however they wish. The astrophotographer may not want to have spent precious exposure time on an image doomed to be discarded at the session's end.
If PHD2's RMS values were available other than via the user interface (UI), the user could perhaps implement a script to abort that exposure based on criteria they choose; they could abandon that target entirely; they could perhaps simply change the filter selection, or any number of other things. The point is that they would have the freedom to decide how to respond at the time rather than finding out after the session is over, that the image(s) are unusable. I think it is a mistake to presume that if the RA and DEC RMS values are discrepant, the only possible cause is a mechanical issue with the system or bad choices of parameter settings in PHD2 and therefore the user needs to fix those problems first. While that may be true some of the time, I think it is not true all of the time.
The value I see in the OP's suggestion is that it gives the user the freedom to choose how to respond during an imaging session. I think that freedom is a good thing.
The guide actions are, of course, written to the Guide Log, but there is no annotation describing what the numbers there mean and, in any case, they are not the RMS values which represent a moving, historical summary of the guiding. PHD2 is obviously performing calculations based on those guiding actions and providing the results to the UI. If those values were exposed and accessible then, the user would at least have a chance to do something with them if they choose before finding out after the fact that the image is spoiled.
I will also use this topic to bring up another possible future direction which may never be implemented in PHD2 but could have great benefit. That is, if PHD2 guiding parameters and actions were exposed and accessible, some enterprising coder may implement an AI tool that could learn for each user's system what settings can achieve the best possible results during any given session. As we all know, users' systems vary greatly as do atmospheric conditions from location to location and night to night. I think such a tool would make astrophotography much more rewarding for many who work with less than the best equipment at less than ideal locations. I see no reason why such a tool is impractical and PHD2 certainly provides the basis for enough functionality to support it, I think.
Regards,
Manning B