No, I've never seen that system
Looks interesting - in the end, it seems similar to what I did, only using an Arduino instead of the Pi.
My software is pretty rudimentary still, and doesn't support zone control valves - although it should be pretty trivial to add supply and return temperature sensors (something I've planned on doing anyway) and control a set of actuated valves. I supply zones just via the manifold's balancing valves, because the load of the different zones is pretty consistent, but I'm still considering adding a mixing valve - regardless, it would be good to measure supply and return; with a bit of math and some logging of how long the water heater runs, it would be fairly easy to calculate the efficiency of the entire system that way
I've not looked at the I2C bridges - the DS18B20 sensors I'm using are Dallas 1-wire, and they work very well with the 1w driver included in the latest Debian from the Raspberry Pi foundation. Because they're handled via a virtual file system, any number of sensors can be hooked up to the same GPIO pins, and they just show up as different directories. Every once in a while I get a failed measurement (it shows up as -0.62 degrees), but that's relatively easy to filter in software.