Repeated Screws_tilt_calculate results fluctuating

Hi,
I have a homemade mk2s/3s clone with SKR 1.4T and LJ12A3-4-Z/BX inductive probe. Without making any changes (i.e. to the bed knobs, z_tilt_adjust etc) running the screws_tilt_calculate seems to produce quite different results (see below)
I know this could be due to many things but does anyone have experience of the most likely candidate? Or are these variations within acceptable tolerances?
Cheers, J

(PS. // probe accuracy results: maximum 0.693750, minimum 0.686250, range 0.007500, average 0.690500, median 0.690312, standard deviation 0.002318)

Screws_tilt_calculate results:

16:26:41
// front right screw : X 190.0, Y -4.0, Z 0.87237 : Adjust → CCW 00:44
16:26:41
// rear right screw : X 190.0, Y 200.0, Z 0.80262 : Adjust → CCW 00:35
16:26:41
// rear left screw : X -15.0, Y 200.0, Z 0.27687 : Adjust → CW 00:28

17:42:43
// front right screw : x=190.0, y=-4.0, z=0.89100 : adjust CCW 00:44
17:42:43
// rear right screw : x=190.0, y=200.0, z=0.66762 : adjust CCW 00:17
17:42:43
// rear left screw : x=-15.0, y=200.0, z=0.20475 : adjust CW 00:39

17:47:33
// front right screw : x=190.0, y=-4.0, z=0.75775 : adjust CCW 00:28
17:47:33
// rear right screw : x=190.0, y=200.0, z=0.27987 : adjust CW 00:29
17:47:33
// rear left screw : x=-15.0, y=200.0, z=0.21575 : adjust CW 00:37