Two columns per session are enough to prevent a month of drifting.
A drill log does not need to be elaborate. Two columns are enough: what you drilled, and what you heard. The first column keeps you honest about the schedule; the second keeps you honest about the attention.
My log has one row per session, not per round. Something like: gated snares, heard three in ten rounds, still confusing them with compressed room mics. That row takes twenty seconds and tells me more than a page of scores. Over a month, the rows show which features resolved and which ones keep coming back.
The log also prevents the classic practice illusion. A week of sessions feels long, but the rows may show only three actual drills, all on the same feature. That is not a week of practice; that is one practice repeated three times. Seeing the truth in two columns is uncomfortable and useful.
Review the log every fourth session. If a feature has not resolved in four rows, change the drill rather than repeating it. If it has resolved, pick the next feature and write the first row. The log is a decision tool, not a diary.
play without a limit makes the log possible because sessions can be short and frequent. A short session with a row beats a long session without one, every time. If you want to see what the rotation draws from before choosing a feature, the song pool is a reasonable place to look first. The log is not a record of effort. It is a map of attention, and attention is the thing that actually changes the ear. Two columns, twenty seconds, four rows before a decision: that is the whole system, and it fits on the back of an envelope. Start the first row today and read it tomorrow. The next row is the only one that matters, and it is twenty seconds away.