[Proj] MGRS notation in extended UTM zones?

[Mikael Rittri]

I have a question on MGRS notation and overlapping zones. 

On maps with the UTM/UPS grids, the displayed grid 
for one zone may extend a bit into an adjacent zone: 

    "Overlap: On large-scale maps and trig lists, 
     the data for each zone, datum, or ellipsoid 
     overlaps the adjacent zone, datum, or ellipsoid 
     a minimum of 40 kilometers. The UTM grid extends
     to 80°30'S and 84°30'N, providing a 30-minute
     overlap with the UPS grid." 
         -- DMA TECHNICAL MANUAL 8358.1
		Datums, Ellipsoids, Grids, and Grid Reference Systems.
		Section 2-6.3.1: Universal Transverse Mercator (UTM) Grid. 

    "Overlap: The UPS grid extends to 83°30'N and 79°30'S, 
     providing a 30-minute overlap with the UTM grid."
		same reference,
		Section 2-6.3.2: Universal Polar Stereographic (UPS) Grid. 
		http://earth-info.nga.mil/GandG/publications/tm8358.1/tr83581b.html#ZZ26

I am not sure how this should be applied to MGRS notation.  
Perhaps not at all?  

For example, grid zone 30X (covering 6°W to 0°E, 
and 72°N to 84°N) contains a 100-km square named WS, 
but it only about 44.3 km wide, because it is clipped 
by the east zone boundary. So, should the MGRS reference

  30XWS5000050000

be considered legal? Is would represent a position 
about 5.7 km into UTM zone 31, but maybe that's okay
due to the 40 km overlap of UTM zones? Or is the overlap 
intended only to make measurements easier on a printed 
map, and not intended for MGRS notation? 

I guess some common sense judgement is required here.
A software system could reject all MGRS references that 
would be the slightest bit outside their proper UTM zone,
and the system would then catch more mistakes, but it might
also reject some slight, innocent trespassing across the 
UTM zone boundaries. 

I just wonder if there is some standard, or recommendation,
about how this error-checking should be done. 

Regards,

Mikael Rittri
Carmenta
Sweden
http://www.carmenta.com