I used the Wireless Audio Module / MD85XXP1
that can be found here http://www.raytac.com/products.php?subid=18
raytac can sell you the module on a small demo board.
I did solder the module to vero board. but its very tricky !!
the pitch is 1.27mm and there are no holes on the edge of the module, just bumps!
I managed to solder 1.27mm headers to the module and bend every other pin so it would fit in double rows to 2.54mm normal prototype board..
to be honest I would just pay the extra 25$ and get the dev board with the modules if your going to test the modules..
I used a CCdebugger from TI , cost 50$ and flashed the modules via the 10 pin header on the demo boards and my own 10 pin header on the module soldered to the vero board.
I would say its not really worth trying to make your own 4 layer cc8530 + cc2950 PA board unless you really want to go into production with something.
there is an example schematic of diversity antenna switching from TI and you can add the Skyworks switch for antenna diversity. does it help with range? a bit, but not that dramtic.. perhaps slightly less dropouts when you walk about with the TX.
you can have a very high quality audio link when you configure cc85xx in PCM24 but you will find latency goes up.. low latency 13ms is only available in SLAC mode.
theres plenty of info on the TI forum and you can get answers to many questions there.
One really odd thing is that if you want to use antenna diversity with more than one microphone is that the diversity antennas need to be on the microphone. Its just the way the Slave - Master topology works out.
Even more funny was I found the audio would get choppy if I uploaded a picture or streamed video over the wifi in my house.. and would completely cut out if I turned on the microwave oven.. must have a leeking microwave :)
Drop me any questions if you want to have a go at building this, would be happy to answer..