if my calculations aren't wrong:
48 of these
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=493-3326-ND
in series would be 2500 farads / 48 in series=52 farad. at a voltage rating of 2.5*48=120 volts.
so 120 volts at 52 farads. if money really isn't an issue you could build that , but the peak current if you shorted it would be insane because each cap has an equivalent series resistance of 4.0 milliohms.
you'd want to be real careful
so it might be better to connect them in parallel for 120,000 farads at 2.5 volts.
if you discharge from 2.5 to 1 volt that's 50 amp hours! That would be an awesome capacitor bank for sure! the price would be around 8k$ though. i think it could put out around 40,000 watts peak too, and according to E=(C*V^2)/2 would store abut 104 watt hours if you could discharge it completely.
These appear to have the same lifetime rating as electrolytic capacitors(2000hrs @ 60C), which can last longer then batteries by far under certain conditions. assuming the same formula for electrolytic capacitors to calculate lifetime(a factor of two improvement for every 10C reduction) that should be 2 years minimum at 30C, or 8 years at 10C. if you were doing something like a high altitude sensor or something where it didn't get to hot (or you insulated the capacitors and had a thermoelectric cooler cooling them) they could last a really long time.
so they are definately useful for energy storage for some uses. one great feature is that being capcitors as long as you stay under the rated voltage you can't over charge them, they just stop drawing power. this makes them great for when you don't want to build charge control circiutry. they also have very good peak current and cycle life (500,000+).
so i would say they are useful for some things but you should always use the right tool for the job.