Gantry twist wierd

Guys, help me solve this problem. Here's a screenshot of the bed frame grid.

And here's the result of the BEACON_MEASURE_GANTRY_TWIST macro.

Gantry twist relative to the center
 High gantry twist: 126.883912μm. 
High chance of first layer problems, beacon scan compensation is highly encouraged.  Front left: -78.091284μm 
Front center: -36.436779μm 
Front right: 16.160590μm 
Left center: -65.383145μm 
Right center: 75.452585μm 
Back left: 4.029200μm 
Back center: 52.003825μm 
Back right: 126.883912μm

How do I know if the macro shows a bump in one corner, but the grid doesn't, and who should I trust?