It looks interesting, although as said above, not quite new in itself.
One thing I'm wondering is whether it would not add some overhead/latency compared to a traditional architecture? Also would it lend itself well to pipelines?
I think (if you haven't already) you should compare this with a classic architecture point by point and see where all the drawbacks and advantages would lie. One clear advantage would be modularity, but are there many others?
Prototyping this on FPGA would be a first step to really get some people interested IMO. While it may be interesting in itself, it's again not very new and cpu concepts are to be found by hundreds out there... so a pure concept is not very likely to attract a lot of interest IMO.
As to comp.arch, I hadn't been there in ages. Whereas many/most usenet groups these days have little activity (or still a lot like comp.lang.c, but with a lot of trolling), this one seems still pretty active and not too polluted. I'll have a look.