Hi, me again. I've done some more troubleshooting and think you are right about the memory on the stereo. I saw this same effect on the battery charger and the inverter...they all show 20 - 50 mA current even when "off". Then I tried lifting the negative return cable from my negative bus for these individual circuits and the current stopped flowing, which I believe rules out a ground leak as the cause for their minor current draws.
I've got things cleared up now so that I am not getting any current flow more than a couple mA when measuring between the positive cables and the battery positive posts.
I picked up Charlie Wing's Boatowner's Illustrated Electrical Handbook which suggested an additional test for current leak: measure DC current between the green bonding conductor and the through-hull or other fitting it is protecting. I did this with the easiest item to measure: the springy grounding conductor that rides on the prop shaft. It is showing 30 mA of current. Wing's book (page 88) says even 1 mA is cause for concern on this. I am seeing this current even when I completely disconnect the batteries (remove all positive and negative cables from batter posts). Is this a real cause for concern? Where could it be getting its current? It looks like it is flowing from the bonding wire to the shaft (red/+ multimeter probe is on the bonding conductor, black/common probe is on the shaft).