Perhaps just because I can't find the relevant documents; then my question is where can these be found? How can it be ensured that future users are better informed than i currently am.
You're right, a lot of these policies were just known between the moderators, and was recently put in written form in the form of a moderator manual (which, also was only decided within the moderators). I'd be happy to make any details public, but not the entire manual, because well, it's a bit harshly written in some places. and here's the moderator manual in full.
Where is it defined what enough means?
Yes - I noticed that this isn't clear, and thought of editing the detail in, but decided against it because it's only last year's nomination thread. "Enough" like on all other community moderation threads, means that the people in agreement must outnumber the people in disagreement by 3, which usually means a score of "+2" (but is "+3" if the nominator retracts his vote, and "+4" if the nominator retracts his vote and downvotes it instead, because the nominator can't actually vote on his own post).
Additionally, the nominee must accept their own nomination (they need to be alerted about it first) and must have at least 500 reputation points. Administrators need to be approved by polarkernel.
Is putting [ELECTED] behind the answer the confirmation of being elected? What if afterwards comes a downvote?
While the previous election was a bit messy, "[ELECTED]" is placed only after the full duration of the election, which is generally 1 week, unless
Why was the thread used in 2015 (against the title of the thread) to elect another moderator according to vague rules?
That was anomalous, but necessary to replace Ron (who had stepped down).
Is it an official impeachment process?
Does a simple majority suffice for it becoming law?
No, it's necessary to have an edge of +2 for the team trying to expel the moderator. There was, in Dilaton's case, during the 7-day timespan of the moderator review - 3 people voted for the expulsion and 1 (me) against (you can't count Dilaton's own vote).
How are the results of the different questions combined to an official outcome?
There's only one question - should Moderatorname step down, or stay?
Where is the official outcome permanently recorded for everyone to see and to check?
It should be recorded in the moderator review itself - I admit the particular moderator review was hasty and unsystematic, I was also hesistant to edit the question body, because drake (who first submitted the moderator review) was for the stepping down, while I was against it, so I didn't want to be accused of manipulating the content of the review to my political standing.
How is it determined whether a new moderator is needed? Or when the next election should take place?
Generally, moderator elections are held on 4 April of every year (the anniversary of PhysicsOverflow's public beta), and administrator (currently known as "super-administrator") elections (which are not as democratic, because polarkernel's approval is needed for any elected candidate, and rightfully so) on 4 December of every year, on the date of Ron's 1-month suspension from Stackexchange, when it was first clear that a site like PhysicsOverflow is necessary.
For the next election this must be structured in a clear and professional way.