Just now I continued with my attempts and it seems that I managed to find a reasonable workaround.
my findings:
The MOOD tag in WMP - is written to the PRIV frame with key WM/MOOD. This is not the regular implementation of thie tag. (See my comment above). Turns out that each letter in that tag is separated by the NUL character. For example the word "dance" would show up in that tag as d(nul)a(nul)n(nul)c(nul)e(nul).
When one attempts to choose a different reencoding as suggested by redwing, the nul is interpreted as a space character, ("d a n c e" in the above example) which is no good.
It would be possible I am sure to write a program that would skip the unnecessary nul characters, which I beleive is the highway solution.
Not ready for this project I chose a roundabout:
i used the only program that I know that deciphers correctly WMP MOOD tag, "tag&replace".
With "tag&replace" I copied all my MP3 file MOOD tags into the COPYRIGHT tag. Following a suggestion in another forum I chose the COPYRIGHT tag because it is a universal TAG that is implemented in most players.
The copy process of "Tag&Rename" not only copied correctly the MOOD tag into the COPYRIGHT tag but surprisingly also into the engMusicMatch_Mood frame. And it was copied correctly, without the unnecessary NULs.
After introducing the MP3 files again into the MB library, not only the COPYRIGHT tag showed correctly the MOOD content, but also MB updated the MOOD tag. It added that same content to the previously garbaged MOOD that was originally incorrectly encoded from the PRIV frame with key WM/MOOD fo the WMP.
In other words MB combined the content of both MOOD tags into the MOOD colomn in MB.
This MOOD tag is serviceable. YOu can just ignore the garbage at the end of each MOOD raw. At first I had in mind to leave that column as is thinking that if I copy the COPYRIGHT column back to the MOOD tag in MB it will overwrite the PRIV frame with key WM/MOOD to blank. If that would happen I would loose the MOOD tag when using WMP. However, luckily apparently this copy process leaves the PRIV frame with key WM/MOOD intact, while getting rid of the garbage characters in MB. So I intend to copy the COPYRIGHT tag into MOOD in MB.
In summary, this is a feasible roundabout way to copy MOOD form WMP to MB. However I think that it would be better to have MB fixed to be able to read WMP MOOD directly.
It is also not clear to me why Tag&replace uses the engMusicMatch_Mood frame, rather than the common TMOD for MOOD, and how MB knew to read and then add the information from two esoteric frames into its MOOD tag.
Perhaps someone here can enlighten all that.