Thanks to everyone for all the info -- it's gratifying to realise that I'm not completely out to lunch, and rather mortifying to find that not only has someone had the temerity to think of it first but the gall to put in place a large-scale development over many years. Oh well, the SLF retirement fund will have to come from somewhere else...
Getting back to the technical stuff, the most interesting aspect (for me) is the nature of the datalink, which is where TCAS is quite weak. There's an awful lot of stuff to wade through, but it doesn't *quite* look as if the system will do the pocket TCAS I was first thinking of -- although the basic time-slice protocol sync'd to GPS derived UTC would work there. But we're eight years out from deployment: plenty of time.
There are some new developments in radio technology which may simplify things considerably, but looking at the timescales for ADS-B I wouldn't think that they'll be mature enough for consideration for this cut of the standard.
Back to wading through the docs. Perhaps I should become an aviation correspondent and get paid for this stuff
R