Well hopefully something eventually comes of this topic, no need to rush obviously. One problematic aspect is that different speedrunners have their own idea of what makes a map interesting/worthwhile to record on. Take slayer for example, the maps seem very linear with high monster density and moderate difficulty (very characteristic of Richard Wiles maps). It looks to me like UV-Speed would basically be Max route while trying to minimize killing and secret detours for the majority of maps. I think a megawad chosen for Compet-N should seek more variety in the maps. While naturally, we shouldn't assume too much in choosing the maps (a few "impossible" pacifist or nightmare situations could lead to nice surprises down the line), there is considerable value in finding a mapset with a good variety of gameplay situations and route possibilities. If we were to choose an already popular megawad, it might be useful for players to discuss potential problems with the wads or changes that could be made, and then work as a group to release a Compet-N edition (I believe the authors should be contacted about this even if the textfile says its ok). Just throwing that idea out there.
Scythe honestly would fit CN well, better than several wads already in CN imo. To circumvent the problem brought on by prboom demos, perhaps there could be a 6 month period where only improvements were considered, unless it was by the same author of the original run. For example, I would have 6 months to try to equal my prboom 0:26 Scythe map 1 max record, and if I failed to get my own time again within the 6 months, Radek's vanilla 0:27 demo would be taken as the record and anyone can do 0:26 for a CN point. The only thing I don't like is having the huge slaughter of map 30 be at the end, as it significantly lowers the likelihood of seeing an optimized 30 max movie. Since most of the prboom players with records in Scythe are still contactable (Angus and Psichotik the only ones I'm unsure of), it would probably be helpful to get their opinions on this as well.
I haven't played TVR much, but I've seen it suggested by several experienced players, so I think it is worth considering. My only gripe with it is that there appear to be only short maxes. Very few current maxes exceed 3 minutes, and none exceed 4. I think it would be more interesting to have closer to a a Doom 2 level of variance in map length, perhaps with a few longer maps.
As for Hell Revealed 2, I don't think it is very viable for categories besides max. Yes, there are some maps which are, but the vast majority of the maps are unsuitable for UV-Speed (big problem), and absurdly hard for nightmare/pacifist/tyson (not as much of a problem, but still should try to be avoided).
Or we could try to find a less demoed megawad. Altima mentioned Doom the Way Id did, whose prboom demos are all improvable. There is also Doom the Way Id did II coming out, which might be a better candidate (DTWID mostly has demos done by only two people, so perhaps it is too specialized).
One final idea I've entertained is that we (meaning those interested basically) design a megawad specifically for Compet-N. I don't think it would be too difficult to garner interest for such an undertaking, and this would insure that gameplay is reasonable for multiple category recording, that there is variance in difficulty, length, and gameplay from map-to-map, and that there aren't any "too easily optimized" situations such as Scythe map 7 where equaling the pacifist/speed/nightmare record is trivial. Naturally, this would be the most involved option. There would have to be lots of playtesting, mappers willing to make big changes for the good of the wad as a whole, etc.
In general, although the choice of wad should not be made lightly, I think that most choices would see reasonable amounts of competition and activity, but I also think it would be best if no new pwad was added until CNDoom is updated to the extent that it is allowed for Compet-N. The combination of the two could effectively cause a much-needed surge in CN submissions.
Just some random thoughts, would be cool to see more discussion on this.