What was the rationale for the non-standard UTM zones near Norway?

The closest I've found to an answer is in the DMA technical manual 8358.1, which states:

3-2.2.1 For that portion of the world where the UTM grid is specified (80° south to 84° north), the UTM grid zone number is the first element of a Military Grid reference. This number sets the zone longitude limits. Zone 32 has been widened to 9° (at the expense of zone 31) between latitudes 56° and 64° to accommodate southwest Norway. Similarly, between 72° and 84°, zones 33 and 35 have been widened to 12° to accommodate Svalbard. To compensate for these 12° wide zones, zones 31 and 37 are widened to 9° and zones 32, 34, and 36 are eliminated.

I emailed USGS and got this response:

Greetings Jaime: zone 32 is 9 degrees wide in zone designator V is 9 degrees wide for the purpose of including the west coast of Norway within a single zone. Zones 33 and 35 are 12 degrees wide in zone designator X for the purpose of putting Svalbard into two zones instead of three. In both examples, the UTM system was modified for the purpose of convenience.

I found an article[1] about the history of grids and datums in Norway. I do not understand many of the details in the article, but it seems like 32V was extended for compatibility with previous grids. The article does not offer a clear explanation about Svalbard.

[1]: Mugnier, C. I. Grids and Datums: The Kingdom of Norway. Photogrammetric Engineering and Remote Sensing. American Society for Photogrammetry and Remote Sensing (ASPRS), October 1999. pp. 1129-1132. Accessed online on 20 August 2011. http://www.asprs.org/a/resources/grids/10-99-norway.pdf.