Sounds like I may have the same problem.
Last session I right clicked an image to “Center Here” and it successfully plate solved but didn’t get close enough for acceptance. I think it may have done this twice. On the next attempt it went into Spiral Search and failed. I noticed that the FOV setings were correct in the first two attempts (104x84) but wildly wrong in the final one pictured (209x167). I have no idea where these values came from. I should add that it usually fails on the first attempt with an erroneous FOV. I have searched high and low within SGP to find these erroneous values without success.
I have had “Center Here” working in the past but not for quite some time.
All other plate solving seems to work, including the Solve and Sync, Blind Sync and the “Center Now” option in the Target dialogue.
It seems to me that for some reason the “Center Here” process is getting erroneous FOV data from somewhere.
UPDATE. I have just gone back into my PlateSove 2 settings and the erroneous FOV values had overwritten the good values I had entered previously, and which had been used by the 1st two plate solve attempts. Very weird - where are these values coming from?