Personally I dislike this in the 2012/2013/2014 year rules, because the contest explicitly trying to "control" my distribution rights.
For example, when I published Zombie Near, obviously it "stays" within the contest archives. For cartridge publication I was being lazy, but publisher (Matra) asked me for an enhanced version and they were right, specially in MSX scene, if people thinks it's same they simply download the ROM and don't buy the game...
So if you publish a game in the MSXdev under these new rules, it limits indirectly the publication of homebrew cartridges for MSX.
I've commented about this before http://karoshi.auic....pic,2281.0.html without official response from organizer, only the previous contest maintainer said it shouldn't be read strictly.
Anyway it was enough to keep me away from the contest.
but from other side, the purpose of these game development contests is, not only for keeping the scene alive, providing more new games easily accessible, incentive people to learn game coding, having fun and joy during the contest, sharing knowledge and experience, and so on
for example, what i loved in contests like CSSCGC is that we see a lot of ludicrously stupid and funny stuff there, and there are no prizes at all in that contest - in 3 years, i think i submitted more than 5 games there! (and that was what motivated me to learn Boriel’s ZX-Basic Compiler, and found a huge potential from that cross-development tool to code to all other z80-based hardware, as it was for msx, sg1000, master system (haroldoop also uses Boriel’s for that), gamegear, etc. )
the problem is, when you restrict the access to the resulting roms, people would be also restricted to know what the game is about (unless people share videos of the game at youtube or so on...), and the purpose of the contest becomes totally nonsensical... :S
and personally, for me, for my personal case, is essential, not only sharing the roms, as i think it's even more important sharing the sources (which is what i'm about to share next week, at msxdev'14)
and from me, everything i develop is purely for joy, experimental stuff, and having the development a part of the experimentation, learning and knowledge/experience sharing process
i wouldn’t see as a problem if the "publishing" process could be only complementary to all of this, but it is when it hugely restricts my freedom and pleasure of doing this