Up front, all of the below is very speculative, and Denkt's likely to provide a far more useful answer, but it wouldn't any fun not to try
.Even assuming no bonuses, I expect that converting military to civilian may not help as while our 'combat' equipment will date, we could be using it to build trucks or the like that would still be useful when war breaks out, and we'll need infantry equipment for training to boot.
Going on the numbers in the OP (which may be subject to change), assuming no bonuses (which may also change), and no benefit in the produce of the military factory early (which I'd think would only be the case for the US - everyone else is likely to be at war soon enough that even 1936 equipment is better than none at all), then for 15 civilian factories, it takes 96 days to create another civilian factory (Option A), and 53 days to convert a military factory to a civilian factory (Option B).
However, at some stage in the future (assuming we're not playing a pacifist Germany and don't think anyone will declare war on us
) we'll need military factories again, and in Option A, we go from 15 civ/1 military to 16 civ, while in Option B we go from 15 civ/1 military to 16 civ/1 military - so when war comes, while we'll need to convert our new civ factory to military (another 26 days - but assuming we don't start converting our new civilian factory into a military one right away, between construction and conversion it will be usefully building more things for us) we'll have more factories overall (so the cap on number of factories per province starts to come into play as well).There's also the question of whether we can use factories in the process of being converted (ie, is it a binary civilian or military state, or is there a 'being converted' state where they don't produce anything at all?)
So it depends in no small part on the payoff we get from the civilian factory as a civilian factory that's been converted - without bonuses, that's around 800 days in-game, so a bit over two years, and another 400 days if we want to convert it back to a military factory again (so closer to four years, although the 400 days is earnt back in military production, not civilian).
That said, those bonuses for production between different mobilisation states are important, and then there are ministers that can help things along as well, and I'm sure some of the techs will change the equation yet again. Pile on some bonuses, and that useful turn-around could go from around four years to around two, and it becomes far more rational (if a tad fiddly - I doubt I'll ever play it this way myself, I'm not really a min-maxxer) to convert military to civ and back again, assuming (which may not be the case) that there's limited value from what those military factories are producing (ie, while the tanks the US may build between 1936-1940 - M2, for example - may not be any use in combat - when the US gears up, it's going to need a lot of equipment for training, so all that old stuff will still have a purpose).
It feels a little 'off' to me that it would make sense for any nation, even one that didn't expect to end up at war, to be drawing down its military production capacity in 1936 (or later). If the old equipment wasn't useful enough for training (which I think it probably will be) it makes me wonder whether there's some value in gaining some experience (land, air or sea, depending on what's being produced) from producing things as well. That would make the decision to convert from military to civilian a far less sensible one (although possibly still sensible in some circumstances), as the US could then use this experience to design variants, or redesign divisions, in ways that were plausible and sensible. This last paragraph is just the historical plausibility person in me rambling though, I'm not suggesting changes to the core design (although I do think there could be value in experience from production more broadly, although am just thinking out loud - need to actually play the game and get a feel for it first to really have any idea
).
