One of the things that will make a vast difference to range, both for ADS-B and for PAW, but especially for the latter, is antenna location. For those of us flying all metal aircraft this has to be an important consideration. For example, the obvious place, on the instrument coaming will mean that signals will be profoundly attenuated below and behind the aircraft. This is not so much a problem for ADS-B, where we are only receiving but for PAW that means that a following aircraft won't see our transmissions and nor will we see his.
This got me thinking about where to put the antennas. From an engineering perspective it is quite easy to have the antennas wherever you want them, either on coaxial cable extension leads or by remoting the dongle with a USB extension cable. If we are flying a PtF aircraft then in principle we can put antennas wherever we like. More difficult for C of A aircraft unfortunately.
For ADS-B reception it is probably fair to say that most stuff will be above us, at least for the time being. As more GA aircraft get ADS-B capability this will change. ADS-B transmissions from the aircraft transponder are at considerably higher power, in the order of 200W peak power typically, so we can afford some losses. An antenna on the fuselage roof should be sufficient for good all round coverage and it might even be that one on the coaming will be adequate. As an aside, transponder antennas are always mounted on the aircraft underbelly because they are transmitting to and receiving from the ground-based radar. We, of course, are listening to the aircraft's interrogation replies rather than the ground radar.
For PAW transmission/reception things are more tricky. The transmit power is quite low so we cannot afford much attenuation. An antenna on the roof and another on the underbelly coupled to the ARF via a coaxial power splitter cable would do the trick but is awkward to engineer and probably somewhat OTT. It's definitely something we need to think about to gain maximum benefit from PAW.
Finally, I'd like to open up the Pandora's box of RF interoperability. We have an ADS-B receiver operating on 1090MHz and a PAW transmitting on 868MHz. It's a fair bet that the front end of the ADS-B dongle isn't the most selective of receivers and a nearby transmitter, albeit 200MHz away, might well swamp the ADS-B dongle's front end. The general rule of thumb here is to keep the antennas out of each other's near field, usually considered to be 2-3 wavelengths. At 1000MHz the wavelength is 30cm, so ideally the ADS-B and PAW antennas should be at least a metre apart to be on the safe side. It would be fairly easy to do some tests to see whether this is a real problem or not.
All this is a bit theoretical but it's based on decades of RF engineering work as well. As we gain real life experience we'll get a better handle on the performance issues and how to fix them. It would be interesting to hear other pilots' thoughts and experiences.