While this doesn't necessarily explain some of the oddities @beanie42 mentioned, this is what we know so far about it:
Pure points is too simple a system, since a guild in Bracket say 167 could stomp all the other guilds in their brackets who decide to not post defense teams because their not interested in Guild Wars, giving out first guild a perfect run and score, which could catapult them up to Bracket 1.
My guess is that avarage point gain per day, average wins as well as total points and your placing in your bracket are part of the formula, and are compared to guilds both above and below you before deciding who moves where.
My wild guess here is that Match Masters were more consistent with their scores, and managed a better average daily score, while the other guilds had some very good days, and some not so good days. Again this is completely my own conjecture as I have no way of knowing how any of those guilds did.
Also I don't think we'll get a full answer on promotion/relegation, as it would allow undesired play methods to occur. Eg Say you need to only be ahead of the team at the bottom of the bracket above you, and you need to be above 500,000 points as a guild. And say you get some very easy matches early in the week, and the bottom guild in the bracket above you has an abysmal time for some reason, you could in theory see the first guild deciding to stop bothering with guild wars and just not put up much of a fight for the rest of the week, giving their later opponents a (while easy) dis-satisfying match. This is similar to what you see in competitions like the World Cup where teams that win both their first matches in the group stages and are guaranteed to go through (due to other early results) will field second-rate teams for their third match to save their better players for the knockout stage.
In conclusion, I'm not saying more transparency isn't need it, I think it is. However I think it's more likely that we'll get "factors" that influence the decision rather than an exact formula.