About to try Marlin1.1.0RC4...

[lakester]

Hmm (again). In ROBOCOP's case, the G28 X0 Y0, if that is the actual gcode being used, should not create any z-movement at all.

MBL is disabled during most homing moves, but even that should produce no Z movement due to the loss of a previous correction. MBL IS enabled when G28 implies or specifies a Z homing operation (though not during any XY homing moves with which it is possibly combined).

For G28 operations that imply or explicitly specify Z, then in MOST cases, the Z move will be done last, after any XY moves (which is opposite the behavior ROBOCOP describes).

FWIW..., G28 X0 Y0 is one of a couple special cases for homing. It MIGHT be worth trying uncommenting the line:

//#define QUICK_HOME //if this is defined, if both x and y are to be homed, a diagonal move will be performed initially.

in Configuration_adv.h

..., though I kinda doubt this will inhibit the z-move you're seeing..., since I'm thinking the G28 may NOT be the culprit.

Anyways..., 0.02USD...

Note: almost forgot: the MESH_HOME_SEARCH_Z is MOSTLY involved with the bed leveling procedure itself..., I don't think it actually is influencing anything seen during "print time".

[lakester]

Small update:

I paid attention to z-movement at the end of the past couple of prints.

There is no z-movement during the end-of-print homing operation. FWIW, I'm homing XY with separate G28 commands, i.e.,
G28 X0
G28 Y0

When I print next, I'll try it with G28 X0 Y0.