Possible canned cycle bug?

Hi, while I would like to try the new G71 coding, right now Ive been using standard canned cycles for simple basic roughing or just programming  it all offline, that being said , I have this code for my jaw boring routine.
%
(BORE JAWS)
G20 G7 G18
G28 G91 X0
T6 M6
G95
G55 G43 G90
G97 S1462 M4
G0 X2.0 Z0.2M8
G96 D2500 S550 M4
G17X2.0Z.2
Z.03
G81G99R.03Z-.49F.01
G91X.1L89
G80G90
G18
G0X10.95
Z.05
G1Z-.5F.003
X1.75
G0Z.5M9
M5G28 G91 X0
Z6.
G90G49
M30

I use a G81 cycle with an incremental step to ruff diameters, G83 also works well if you have stringy material that  doesnt like to chip break, works on OD/ID. 
Issue:  when using an incremental step for each pass, the cycle doesn’t  cut to depth, in the above code, the final z depth should be -.49, when it in fact only cuts .43 deep. It seems as though lcnc is adding the .030 R distance twice and subtracting it from the final z depth. to get a depth of .49, Z had to be programmed to -.55. Is this a bug? Did I miss something? Ive been programming Fanuc type controls since 1991, never ran into this one. Suggestions? Thanks

 

I should probably add that this issue appears only when the step is in incremental , G91X.1L50. watching the dro, I want to say the dro says .43, even if the Z is programmed to .49. Ill make a short video tomrw.

In absolute, the z depths function as expected.

example

X2.Z.2
G90
G81Z-.5R.03F.01
X2.2
X2.3
X2.4
ect, will bore down .5 as would be expected.