Please see bellow image from 5G measurements.
Do you see anything wrong related to serving and detected beams and PCIs?
This picture is from NSguru handset tool and measurements are updated each second or so on the main screen.
What I find it to be odd: first of all serving PCI + beam 157 + 4 is weaker than detected PCI + beam 157 + 3. This is not normal.
Second issue there is lot of pollution i.e. too many PCIs within 5 dB window of RSRP (157,156,220) which clearly means an area not optimized.