If you decide to start documenting stuff the key IMO is to make sure you're only writing what you'll actually need tomorrow, rather than what you're thinking about today. Otherwise you can tend to find yourself turning the game into work and putting yourself into a nasty double-bind where you're frustrated when you're being "good" - because writing all this stuff down just feels like work - and just as frustrated when you're being "bad" - because you know you "should" be writing stuff down, but you just can't be bothered.
For me, I never find I need reminders about the tactical/operational side because that's what I spend most of my time thinking about and what is, after all, the most fun. So I know that I'm invading Tarakan tomorrow, I know what the bombardment TF and carrier TF will be doing, I remember that the last hostile naval force retreated back to Java, so I don't write any of that down. It would just be make-work.
The first thing I do keep is a daily to-do list for the next turn, a list of little micromanagemnt or admin tasks that I know need to be done but can't be done this turn. For example an air unit rebasing that will take two hops via some island in the middle of nowhere that I never normally look at; it would be very easy the next day to forget where the unit was and fail to complete its move. Or I find a base that needs resupply, but the port at the supply hub is full this turn. So I write down a reminder to put together the supply convoy tomorrow. Also, when I'm tired of dealing with the more boring logistical stuff - say I really don't want to spend ten minutes collecting shipping for another Port Arthur resource convoy - then I can add it to tomorrow's list rather than forcing myself today. (My daily reminder list is also handy when I've already finished the turn and sent it off, but suddenly have a new idea or remember something that I should have done. It goes on the list to deal with tomorrow.). This sort of to-do list is really easy to keep - one short sentence per to-do - and leaves me with a nice satisfied feeling when I've completed all of today's tasks.
The second thing I document is the really long term strategic decision-making, the sort of stuff that I'll come back to in six months (real) time and probably wonder why the hell I did what I did. So I document all my changes to industry with the reason for each change/expansion; that way future self knows what prior self was thinking and planning. I do the same with decisions about ship conversions and so on.
Here's an example from my aircraft production planning.
Army fighter production
long term planning:
Nate isn't much worse than Oscar I (lower cal MG mainly), Oscar's key advantage is range.
Oscars never have cannon but do have decent range and armor by 2/43.
Tojo mostly don't have cannon but better MGs and performance than Oscar from 6/42.
Tony develop cannon and armor but performance and range are poor.
Frank has cannon, armor good performance and range from 1/44
I like the look of the Nick, potential for ground attack and bomber killer (range, armor, cannon). Not sure how important.
Therefore:
Operate mixture of Oscar and Nate until 6/42
Convert to Tojo as soon as available.
Abandon Tony development
Abandon Kawasaki Ha-60 development (only used by Tony)
Accelerate Frank development as much as possible and convert late 43
20/12/41 Addendum: Only the army has the pilot training pool to supply a large land based fighter and bomber force long term. I was deceived by the number of Zero units at Phillipines and Malaya at start. Navy does not have the training tail to support these. The army is fighting these first campaigns with Nates and Ann/Mary/Sonia, and urgently needs a rapid buildup of modern types if it's going to be the front-line land-based service.
Stopped engine production Nakajima Kotobuki
Stopped engine research for Kawasaki Ha-60
Continue production of Nate until Kotobuki pool is empty (approx 4/42)
Convert Nagasaki Oscar Ia factory to 1c and expand to 30/month
Stop production and repair of Harbin Oscar 1a factory
Convert all Tony research factories to Ki-84a Frank research and expand to 15 each (for a total of 100 Frank research)
Expand Ki-84b Frank research from 15 to 25
21/12/41 - Convert Pete factory to produce another 30 Oscar 1c.
Obviously that's a lot more documentation then my one-sentence daily to-dos, but it's the sort of stuff that you only write once per game but go back to and review quite often (even though I'm only a couple of weeks into this game)
The third thing I document is my operational planning, and here I'm using the process of writing plans down as a tool to clarify my thinking. It's easy to think vaguely "ooh, it's about time I invaded Anbon", and often easy enough to wing it in game pulling together the forces needed from wherever they are nearby. But at some point I find myself realising that I have two invasions landing today, and my one bombardment TF was supposed to be supporting both of them. Or that the ships I sent off in a supply convoy four turns ago had originally been left in that port to transport the troops that have just arrived by rail, if only I'd remembered. So I try to think through and write up a brief plan for each major operation. Here's an example.
16/12/41 Tarakan - Balikpapan: Take both ports in succession, suppressing Dutch naval forces in the area.
Base: Cam Ranh Bay.
Forces: 2 SNLF Bns+ each for Tarakan (loading at Samah, 4 days) and Balikpapan (loading CRB and Samah, 4 days).
Shipping (CRB, in place) (Samah, in place)
ASW TF covering amphib TFs (in place)
BB SCTF patrolling/bombarding (in place)
search from Phillipines CVTF (in place)
Replenishment for CVTF & SCTF (Babaldaob, 1 day)
Opstart: 20/12/41. arrive Tarakan D+6, arrive Balikpapan D+9, operation complete D+12
Again, almost no detail on the tactical/operational side - I can work all that out in game. And I don't feel the need to list all the ships that will be part of any task force, just the role it has to carry out. But by thinking about what forces I need I'm forced to find out where I'll get them from, how long it will take to put them in place, and therefore how long it will be before I can start the operation. By working out how long those forces will be committed to the op I then enable the parallel planning of other ops. So in a couple of days' time when I'm planning my invasion of Anbon, I know from skimming through my existing op plans that the one BB SCTF and replenishment TF in the DEI are committed elsewhere from 20th Dec to 1st Jan, and Anbon will either have to wait until 1st Jan or do without those assets.
Ultimately, I do all this because it enhances the fun of the game (hard to believe though that might be from reading this mega-post!). The reward for the extra time and effort put into planning and documentation comes from the feeling that I'm doing things right, of a plan smoothly coming together in game. If it (or any part of it) started to feel like work then I'd stop doing it.