Here's my attempt at an analytic view of the map dungeons.

What we know about them, from the live letter:
-A portal into the dungeon has a 50% chance of spawning after beating the normal map spawn
-The dungeon contains 7 rooms, each with 2 doors (except for the last room, presumably)
-After clearing a room, you have to choose a door, and choosing the correct one will let you proceed

Now, we can also make an inference based on that information:
-With two doors out of a room, you have a 50% chance of selecting the correct one

This seems obvious, but it could be implemented differently--e.g., rather than one door per room being designated at instance creation time as correct and one as incorrect, "correctness" could just be a dice roll when someone interacts with a door, which would allow manipulation of the probability. That said, doing it that way seems more complex than necessary, so for the purposes of this post, let's assume the straightforward approach.

The implications of the above are, for any given map (assuming your party is capable of clearing each room):
-You have a 50% chance of reaching the first room
-You have a 25% chance of reaching the second room
-You have a 12.5% chance of reaching the third room
-You have a 6.75% chance of reaching the fourth room
-You have a 3.13% chance of reaching the fifth room
-You have a 1.56% chance of reaching the sixth room
-You have a 0.78% chance of reaching the final room

Now, let's put that into more concrete terms. Since the live letter mentioned that it would be best to go with at least a light party for these dungeons, we'll assume that you and three friends run four maps a day (one for each of you), every day for a year; that's 1460 maps in total. Out of those maps:
-You'll get into the dungeon roughly 730 times (twice a day)
-You'll get at least as far as the fifth room roughly 46 times (not quite once a week)
-You'll get to the final room roughly 11-12 times (about once a month)

To be perfectly honest, those numbers don't sound terrible at first glance, especially given the rewards that were mentioned in the live letter. However, it's important to remember the assumption that was made: four maps a day for a year straight. If we change that to be more realistic--say, 16 maps a week (four people meeting up on the weekends to do maps, each with two maps on retainers, one held in inventory, and one used), missing a week here and there--the numbers drop proportionately. A group like this would likely only see the final room 5 or 6 times in a whole year, and a non-negligible number of such groups (about 1 in 300) wouldn't see the final room at all in a year.

Is that reasonable from a rewards standpoint? Maybe, depending on the exact breakdown (keeping in mind that we don't know the drop rate on things like rare minions). Does it seem fun? I personally don't think so, but that is, of course, subjective--and also highly dependent on what the earlier rooms are like. If each room fight is challenging and enjoyable, with decent rewards such that making it into the dungeon at all feels good, even if your group only gets through the first room or two, that would obviously change my opinion to a more positive one.