Thanks for the cite. I've spent some quality time on Mike Holts' forum and, as usual, find the discussions fascinating and the Code in desperate need of work beyond the normal revision cycle.
312(5)(C)(b) immediately rains out the parade by prohibiting penetration of a structural ceiling, which is the reason for using the conduit in the first place. I can't believe that any reasonable person would prefer 30-40 individual cables making their way out of the attic space into individual knockouts on the panel to running them through several large conduits. But, is a garage ceiling "structural"? At least one NEC expert says no. If there's no drywall on the ceiling and the conduits pass between the joists, are they "penetrating" the ceiling? Maybe, maybe not.
312(5)(C)(g) says to use Table 1 of Chapter 9 to determine conduit fill; Note 2 to Table 1 says Table 1 is "not intended to apply to sections of conduit or tubing used to protect exposed wiring from physical damage", which seems to be the case here, since 312.5 starts out by saying "conductors... shall be protected from abrasion...".
Finally, the entire Exception under 312.5(C) applies to a "surface mounted" enclosure. If the service panel is mounted between 2 framing members, is it "surface mounted"? What if the panel is mounted on the surface of a masonry wall, on which a finish wall is later constructed so that the panel is ultimately flush with the finished surface?
Bottom line, as usual, seems to be that the Code is pretty vague and conflicting in may instances; ultimately, if your AHJ likes it, it's OK. There seem to be lots of conflicts in both directions -- one inspector mentioned on Mike's forum, for example, will not accept plastic boxes with the integral cable clamps, even though they're UL approved, Listed, and still have a few drops of holy water on them.