For improved resolution I would suggest increasing NFFT to 65536 or 131072, or even higher if required. At 450kHz IF and NFFT @ 262,144, the resolution bandwidth is reduced to 1.19Hz.
I would have thought Marker B should be spot on if you’ve selected marker B and used the next/prev buttons to move the marker to the required peak. Is this what you have done, but the marker is in error ? I’m pretty sure things will improve once you reduce the resolution bandwidth.
A bug has surfaced with regard to PPM Trim for a connected RSP device. Changing the value may cause the analyser to lock up. In addition, if the analyser tries to launch with a setting file where PPM is not zero, the analyser will not launch. You may have to manually edit the CurrentSettings.ini to SDR_PPM_Correction=0, or delete both Default.ini and CurrentSettings.ini
Let us know how you get on.