Hi VigilantSound, let me try to sum up.
TotalMix is actually in the card, you only use the graphical interface to manipulate routing and mixing. This what makes Zero Latency Monitoring (and ASIO Direct Monitoring) possible - the input is passed directly to some output.
When you use two RME interfaces, there are two possible monitoring scenarios. Either use the DAW software for monitoring (with latency) or create a physical connection between the cards to pass a submix (no latency).
For example: you have 2 cards, A and B. You will monitor all input signal at the output of Card B. In the mixer for Card A, you can create a submix of all your inputs on any output pair, then connect this output to an input of Card B. At this point, you can mix the inputs of Card B with the signal coming from Card A to create your monitoring mix.
Of course, the number of submixes possible is only limited by the channel count of the card, so you can also create several different headphone mixes at Card B as well.
I hope that is a bit more clear.
Regards,
Jeff Petersen
Synthax Inc.