It can be a bit frustrating to have the release delayed time and again. But, what would happen if whatever build is released is too bare-boned? Or too unstable? What good would a 12-turn campaign, before permanent CTD, do? Not much. It could even kill of the interest for the mod. No one likes bad press, least of all the volunteers who spend their precious free time to create EB2. No one is getting paid for their efforts. Releasing it too early can be highly demotivating for everyone involved.

Seeing that a total conversion is quite a complicated affair, there needs to be a lot of centralized planning, a lot of testing to see that introduction of additional features and such do not lead to bugs or other unintended consequences. Sometimes coding can be very annoying - it is not a coincidence that bug-fixing for software usually requires a substantial amount of time. It is even worse when you are creating a mod, since your ability to check for errors and flaws is more limited. Never mind the tools you have at your disposal. It is no coincidence that it took such a long time to discover the "path-length"-issue in EB1 (and other mods).

That being said, it is quite possible that the team has "too high" standards for a first release. Very rarely it happens that a program or mod is perfect as it is, or does not have some highly unbalanced / annoying features. With something like the scope of EB2, it is virtually impossible to avoid, since there will be thousands of lines of code, tens of thousands of lines of descriptions (units, provinces, buildings, etc.), and undoubtedly hundreds of thousands of lines of communication in the internal forums of everyone involved, to ensure historical accuracy, quality, bugfixes, balancing issues, etc..

Bug-hunting and balancing (which will partly be based on the quirks of the combat engine, for example) can be notoriously difficult if you have only a limited number of people being engaged in the task. Especially when there are a few hundred units, and close to 30 factions to check. I am reminded of the AI expansion threads for EB1, in its various incarnations. I think such threads could help substantially with balancing overpowered and underpowered factions, ahistorical expansion tendencies etc. For instance the AI steppe factions were quite weak in the earlier EB builds, due to auto_calc issues.

These are just meant as general observations. EB set the benchmark very high, considering what was possible with the RTW engine. And I am sure that the team for EB2 wants to improve on EB1 in all its aspects, and that is by no means an easy task.