Profil check

Assuming the view from above when it is on the CMM is as per the main view on your screenshot, you would need either an indexing probe head capable of 5° increments or a rotary table and a horizontal probe tip. Align the part as per the drawing - spatial +z would be the datum A cylinder, planar -X would be a line going from datum A to the Ø17.5 hole shown in section A-A, origin would be X=datum A, Y-datum A, Z=point where axis of datum A intersects the top face. In terms of the profile callout, I would definitely do it in a loop where I scan each radius and then either index the probe B angle or the rotary table by 5° each time (360°/72=5°). I would generate the radius using GENCIR with about 180 points to make sure it was a fairly smooth curve and then use ScanOnCurve to measure. Each time through the loop you would either have to rotate your alignment by 5° and then transform your scan back into the main CSY or keep everything in your main CSY and transform your scanpath through the 5° rotation before measuring. You could plot each scan individually using CURVE_P, again within your loop but I suspect the profile callout refers to all of the radii collectively since it is profile of a surface. The easiest way to to do that would be to collect all of your scans into one element and use MESRF but that command needs a CAD model for the nominal data.

I put the part exactly the other way around.
The zero point and the alignment is similar to your description.
Then I rotated my coordinate system by 2.5 degrees and carried out four scans at 90 degrees each with a star button.

I measured the scans without calculation and then fixed the diameter with SETITR.
After that, I calculate the fixed diameter circle and evaluated the shape of the circle

PRG CODE:

USEPRB (NAM=PRB(34))
TRAOBJ (NEW=CSY(1), OLD=CSY_A2, ANG=2.5, AXI=+Z, DTY=CSY, USE=Y)
MOVCMM (TYP=ABS, DST=(215.0,0.0,66.0), CSY=CSY(1))
GENSCACIR (NAM=SCAN(1), PLA=ZX, INO=O, TYP=CIR, XCO=202, YCO=0, ZCO=-6.5, DIA=11, DNS=20.00, SPD=2.0, PDI=3.0, MIP=21.00, MXP=159.00, CSY=CSY(1), RTS=N)
MECIR (NAM=SCAN(1), CSY=CSY(1), MOD=(NOC,NOE))
SETITR (NAM =SCAN(1), TYP =CIR, X=202, Y=0, Z=-6.5, VAL=11, DIR =Y, INO=O, FIX=N, FIY=N, FIZ=N, FID=N, FIA=Y, LCK=Y)
MECIR (NAM=SCAN(1), CSY=CSY(1), MOD=(NOM,NOE))
MOVCMM (TYP=ABS, DST=(215.0,0.0,-14.0), CSY=CSY(1))
MOVCMM (TYP=ABS, DST=(215.0,0.0,66.0), CSY=CSY(1))

...
TRAELE (NEW=SCAN1, TRA=CSY_A, OLD=SCAN(1), TYP=CSY, RPL=N, EVA=N)

CURVE_P (ELE=SCAN1, A_O=XZ, FAC=50.0, PTO=0.025, NTO=-0.025, TYP=1, FRM=Y, SPF=Y)
!Radius ausgeben
SETITR (NAM =SCAN(1), TYP =CIR, X=202, Y=0, Z=-6.5, VAL=11, DIR =Y, INO=O, FIX=N, FIY=N, FIZ=N, FID=N, FIA=N, LCK=Y)
TRAELE (NEW=SCAN1R, TRA=CSY_A, OLD=SCAN(1), TYP=CSY, RPL=N, EVA=N)
MECIR (NAM=SCAN1R, CSY=CSY_A, MOD=NOM)

I had no other idea
the results are very bad .
but i think the parts are so bad....