So, here we are, after understanding that requiring to be defeated was not sustainable, and requiring combat was not compatible with the question in the FAQ, here remains only one possibility:
{more ranks
+ combat + defeated} => {steadfast}
Steadfast not in combatLike in the previous attempts, we need first to check if this is allowed by the BRB.
Let's see p.54.
"Simply put, a unit is considered to be steadfast if it has more ranks than its enemy. "GW does not mention combat, which goes well, but they do mention enemy.
What happens when there is no enemy? Could one consider that the unit has more ranks than zero?
{no enemy} => {zero enemy’s rank}
This can be argued with the help of the sentences immediately after "simply put":
"As with calculating the extra ranks for the purpose of combat resolution, the ranks have to be five or more models wide for the unit to be treated as being steadfast. (…).
Similarly, the enemy’s ranks also have to be five or more models wide to counter your unit from being steadfast.”Here, the mechanism for determining whether a unit has steadfast status can be clearly understood:
1. Check if your unit has eligible ranks.
2. Check if there is an enemy with eligible ranks.
3. You're steadfast, unless the enemy has at least as many ranks.
I detail it more =>
here.What happens with the other sentences in the BRB?
There are all compatible with this understanding.
Even better, the steadfast status of stubborn troops and of troops in building is permanent, too. Very consistent.
Let's read the AB and its FAQ.
The reading "either not in the same combat or in another combat" is understood as "either not in combat or in another combat", i.e. anywhere.
"Always use..." remains under the condition that the FAQ is valid, which necessitates that the AB is valid, i.e. that the parent is steadfast.
With our understanding of steadfast, it creates no difficulty to have parent steadfast when examining detachment combat result.
We see that the only exception mentioned in the FAQ happens when the parent is denied steadfast by ranks (no surprise), in that case it adds one more possibility for steadfast denial against the detachment.
The whole FAQ goes well with our understanding:
{more ranks, even not in combat} => {steadfast} (BRB)
{steadfast parent} => {use parent's ranks for detachment's steadfast} (AB + FAQ, part 1)
{parent denied steadfast per ranks} => {detachment denied steadfast} (FAQ, part 2).
With this understanding of what steadfast is, there remains no difficulty to be solved with the BRB, the AB and the FAQ.
-=-=-
Overall, if one selects this interpretation for being the most compatible with all rules, one should make it clear with his opponent before the game, in case the opponent has a different understanding (this happens).
Just remind him that "either not in the same combat or in another combat"
confirms that a parent unit not in combat can be steadfast.
-=-=-
What remains to be argued is whether parents in building transfer steadfast to their detachment.
They are steadfast, sure, so the AB is relevant. They have zero rank, sure, but is the FAQ relevant?
Here, the argument can be made that the FAQ limits itself to cases when the parent has gained steadfast rules for its ranks, which is not the case in buildings.
But that's a minor argument, independent of the rest of the armumentation.