I always save a set of orders first, then check the move preview, which takes about 2 minutes. If I'm *certain* that's the best option and there aren't any complicating factors that require deeper analysis, I'll hit ready and forget about it.
But, if there are complicating factors on the board (and this is Dip, how often is there clearly one set of "best orders"? Not that often) I may not have time to be sure I have the best set of moves right then - and that might easily be half a day or more away, work and/or sleep and/or a crazy busy weekend or such intervening. Once I get a chance to check properly, then I'll ready.
Same basic situation applies to Ready/Process/Save. It's circumstantial - if I think I need to do more analysis, then I may not even Save.
The general situation is that I suspect a lot of gunboat players are people who are quite busy and don't have the time to devote to full press games. That means they're probably mostly a lot like that - it's easy to grab a few minutes here and there out of a busy life, but not necessarily the full 15-20 minutes to really check out the whole board and do a deep analysis taking into account the whole game history, etc, which is sometimes very necessary, if you're playing to win.
And it's not trivial to work out even "simple" moves sometimes - even apparently simple "only one province" retreats may not be so simple because of the option to disband instead of retreat. (The only exception really is forced-disbands, for which the true solution is to fix the adjudicator to automatically apply them, which is in the feature queue but not likely to happen soon because it's a high risk fix and developer time is very limited.)
Ultimately, the phase length is there for a reason - that's the (maximum, but still reasonable) expectation for how long anyone can take to give orders for any reason.
If you want a quicker game, join one with shorter phase lengths, basically.