Does Phil perhaps have a lower impedance (i.e. heavier gauge and/or shorter) connection between the two batteries?
If the electronics are looking to determine when the battery is more or less recharged to lower the system voltage, a good low-impedance connection woudl be essential to getting the house battery to charge most or all of the way in a reasonable amount of time. Otherwise, the voltage drop in the interconnection will allow the chassis battery to work its way up to what the ECM thinks is "full" well before the house battery, and the charge rate then drops.
It might be helpful to measure the voltage drop under charging conditions, either directly or indirectly (by simultaneously monitoring both battery votlages). If there's little charge current flowing over the connection, there will of course be little voltage drop per Ohm's law.
I could be all wrong, too, and maybe there's some other difference.