Six seasons in and my rule is one new song per year, anything past that is a bonus. Three or four songs ready before the season starts is genuinely normal for most of us doing this as a hobby. The people with ten fully sequenced tracks either do this full time or they've been building that library for a decade. You're not behind, you're just running at the typical pace.
As for scheduling, what finally worked for me was tying sequencing to the physical build instead of trying to do it in seprate blocks. Once the last prop goes up, that's it, I stop touching hardware. Sequencing only happens after dark, one hour a night, three or four nights a week. An hour is short enough that I don't burn out, and it forces me to work in chunks instead of endless fiddling. I also pick the song first and listen to it on the drive to work for a week before I ever open the editor. By the time I sit down I already know where the hits are going to land, and the actual sequencing goes way faster. The trap you're falling into with the old songs is real. Tweak mode is where sequencing time goes to die. I made myself a hard rule that a finished sequence is frozen once the season starts. If somethig looks off while it's playing, I write it on a notepad and fix it in January when there's no pressure. Ninety percent of those notes never get acted on becuase nobody else notices. Also, don't sleep on sequencing only the big sections of a new song. If I map out the intro, two or three choruses, and the ending, I can get something listenable in three sessions. Fill in the middle stuff later or never, honestly. A song that's sequenced loosely beats a song that doesn't exist. One more thing for the neighbors memorizing the tracks: shuffle the order every week and throw in a sequence you already had that you never actually ran. Same three songs feels fresh when the order changes.