since I am using in/out AND out/in on both G5's the clock of the Metric Halo is being recieved by the slave MAC.
Whether that's really what happens is an interesting question. I wouldn't assume that it is, but who knows. It might be an asynchronous connection, i.e. the Mac might convert the SR. Not important, though.
Ed, whether you hear a difference in clocking set-ups depends entirely on the specific equipment - whether it has a PLL or double PLL, etc. If you listen to Bob Katz (no, I'm not one of those people who points everyone to his website to end every discussion about digital audio, although he is obviously not just some casual guy [:)] ) there's something wrong with anything that sounds better with external clock and you should complain to the manufacturer; a clock itself doesn't have a sound of its own.
However, I can say beyond a doubt that when I was using the Digidesign 888 and 888/24, using *any* external clock made a noticeable improvement - regardless of whether the clock came from a cheap box. And when i was using a digital mixer in my rig, the best-sounding set-up was when I used distributed clock, again regardless of the box.
Likewise, a good friend of mine who uses MOTU gear says the Big Ben makes a difference even his untrained wife can hear immediately. So I guess it all depends.
In any case, I can hear the difference on some material when I use the embedded clock in a lightpipe, especially over longer runs (> 3 ft. or so). But since jitter supposedly only affects converters (again, I know that's not 100% true), meaning the transfer either succeeds or fails, it's only while you're monitoring that you can hear the jitter. Supposedly.
Note that jitter is just a subtlety, while clicks and pops are a sign that something is broken. Ed obviously knows that, but...