I kept falling behind, but I'd sit down and make a dozen rooms at once. I did mine for HeroQuest, so I'd sit down and make a quest or 4 in one sitting, then return to them later to fill in the text from my point form notes. I ended up "finishing" on Jan 1, with just one quest needing to be populated and some text choices. All that's left is to copy a big text block a few times for a unique room that shows up in a few different quests, but I have that written down somewhere. I also need to do a little Photoshop work to clean up some limitations from the software I was using.
The lesson I learned was... don't fall behind. As for other learning, while I did have a few quest chains and some minimal planning, I really just kind of slapped down some stuff I thought looked cool or parts I hadn't used yet. No real theme (other than those quest chains), and no consideration made to anything like balance. There are some quests that are narratively the same, and a few repeat mechanics, but I'm otherwise happy with the project. If I did this again, I might plan it out a bit more and make a few cohesive questbooks (campaigns), rather than just random ideas and largely isolated quests.