Well, my take on this (again, just speculation) is that there are two levels of "decommissioning" that could occur. One level is to simply pull out the culverts and remove fill from the stream crossings, allowing the natural drainages to function unimpeded. The next level is to destroy the road surface between stream crossings, and this is where a road can become a trail. The difference, IMO, is that the full decommission removes open drainage ditches along all portions of a road, and begins to restore the natural runoff profile -- and this is probably an important step in prevent road fill failures that lead to slides (from channeled runoff). A decommissioned road probably re-vegetates more quickly, since the surface left behind is a very loose and friendly to plant growth. That's my take on it, anyway.

A good example is the first mile of the Ramona Falls Trail, which used to be road but was converted by basically running a backhoe along one track, and leaving a long pile of boulders and loose soil next to a smooth trail on the intact road track. This is an important distinction to make in commenting on the MHNF proposals, since a decommission contract probably wouldn't be much different, whether the contractor is retaining a portion of the road grade, or not.

The examples I've seen of a conversion generally retain the downslope track in sloped areas, I suppose with the idea that loosening up the road bed in the uphill track is a step toward the cut slope eventually filling in to the narrow trail grade. Because of the relatively gentle grade, these would make excellent bike trails -- I know the mountain bike advocates are watching this pretty closely for that opportunity.

I never thought about the culverts being removed, or the fact that a ripped up road probably would re-vegetate faster than a gravel road.  If these do in fact get ripped up, it would be nice to leave a trail track in them that was somewhat even rather than the very uneven surface of the other decomissioned roads I've hiked.