It's better to have hot air exit where the dynamic pressure is highest, at least...if your aim is to generate vortecies. The pressure underneath the car is lower than over it. In the RB7's case, the area on top of the beam wing has obviously higher pressure than the underside, provided that appendage actually functions as a downforce generating device. It's why the exhaust gasses aren't directly blown at the diffuser, but rather at it's edges. What you want to do is introduce the higher temperatures directly where the pressure differential is greatest, at least that's partly how tornadoes and hurricanes form in nature. Remember good ol PV=nRT, there is a clear relationship between temperature and pressure of a gas.
There is also electricity and magnetism involved in vortex generation but that's beyond the understanding of mere humans and I doubt anyone could do that with an F1 car currently. It would require the manipulation of a magnetic field, and theoretically you could use the KERS to generate this magnetic field, but how would you control it? Before you disagree with this remember clouds generate lightning by the static electricity caused by the dust in the cloud moving around. Evaporated sea water has some salt in it(ionic compound), clouds need dust not just water vapor to form, and the rain that get's tossed around in rain clouds also has dust which rubs together. Lightning rarely occurs from whispy white clouds because they're stable, there is little pressure differential agitating the contents of said cloud. But you get a cumulus cloud caught in a updraft(caused by a pressure differential(mostly)), and all hell breaks loose.
http://books.google.com/books?id=QznSAA ... es&f=false
Back on topic I think the sidepods are the way they are because of the crash structures, if not I would assume they would run them as low as possible. Especially as battery technology evolves, meaning lighter and smaller battery packs, and better packaging opportunities.