The simple act of writing up stuff in even a semi-formal way is highly useful. For a start you suddenly realise how little you know about something when you take such an approach!
You remember things you had forgotten.
As part of the research you may learn things you were unaware of at the time and may wish to take into account in a fresh observation.
During your education you took notes in order to remember things later for examination, and no doubt like some of us you can still remember some of the formulae and facts you picked up all those many universes ago.
It assists your descriptive and thinking capabilities. Especially when you find you are sometimes repetitive in your descriptions, although things can be much of a muchness.
You may end up using categorisations and classification/hierarchical layouts you had not utilised before which enable you to think and understand more efficiently and/or readily.
You may learn how not to use interminably long compound sentences riddled with clauses that barely anyone else can understand or decode the structure of. I know I never did.
Taken with that outlook even if a final intent of publication is not achieved, whether through abandonment of project part way through or whether for any other reason, the exercise has not been wasteful and in fact has been productive, not to mention somewhat appropriately ideal for cloudy nights.
Whenever I've had to do an amount of research I have come to the conclusion that all that effort, and the temporary files, are going to waste if I don't do something with it before I delete it all as done with, so I have tended to pubilsh it, or email it to someone potentially interested. Or in recent times bung some blather in these fora. It's an attitude of 'well, I've put all this time into this bludy thing, time to do summat with it more than leaving it encoded on electromagnetic domains!' That, of course, only applies to work with a successful outcome. Unsuccesful work doesn't lead anywhere. That is irrelevant to observational logging.
You've gotta wanna though, I can touchtype which helps, but making neat presentable and interpretable by others illustrations, in my case invariably graphs, is tedious and time consuming.
And finally, a recent post somewhere on CN noted that the legendary NSOG volumes 1 and 2 came from what used to be a home published magazine style observing thingy consisting of the observation logs of a couipla folk which were posted out to subscribers before it became a book compendium. Before print on demand, no doubt!
Well, that means there's precedent. Doesn't mean you'll achieve that, but there's precedent and interest.
I don't think you'll outdo Burnham though ; )