When is it acceptable to use a computer in postal chess?

When is it acceptable to use a computer in postal chess? I'm not talking about the obvious method of cheating by entering the current position into a computer for a recommendation of the next move, but I'm talking about more unclear uses of computer for that game where you study possible later positions that might occur. Some examples:

(1) Learning more about variations that *might* occur later in the game, say by getting a feel for the Poisoned Pawn Variation of the Najdorf Sicilian (which doesn't start until the 7th move: http://en.wikipedia.org/wiki/Poisoned_Pawn_Variation) by playing many games with that initial setup after you find your opponent played the Sicilian Defense on the 1st move. If that's acceptable, then how many moves distant from the current position would it be ethical to practice in a possible future variation?

(2) What about typical endgame practice for the opening variation that is being played? For example, the Sicilian Defense usually results in an endgame with Black's 2 queenside pawns against White's 3 queenside pawns, and an asymmetry in the kingside pawn majority, too. So would practicing engames with that pawn balance be ethical?

If it's *not* acceptable to use a computer in any of these capacities, then what if you're about to enter a live tournament at the same time your postal game in going on, and you need to practice the same opening or variation on your computer as you usually do before such live tournaments? Does that mean you have to do worse than you usually would in the live tournament just because you have a postal game going on at the same time?

I think there's nothing wrong in using an engine for the above said things.

Next time try to keep it short, many people just skip the thread looking at such a long question.