As Neil deGrasse Tyson says, NASA was founded on the fear factor of USSR's Sputnik which was launched using the intercontinental ballistic missile technology.
BTW: According to the comments here, the solar power generation on Philae weights at least 20 times less than the nuclear generator would.
Consider that the RTG from the Pioneer 10 mission weighed less than 16kg: http://en.wikipedia.org/wiki/Systems_for_Nuclear_Auxiliary_P...
[0] http://www.ianus.tu-darmstadt.de/media/ianus/pdfs/arbeitspap...
On the p78 is only "The consortium formed of leading institutes of Europe includes the scientific organisations of Germany, Finland, Italy, Hungary, Great Britain and Austria, which have not the technological capacity to manufacture RTGs. It was therefore decided to develop new high-efficiency solar cell for this deep space mission."
"Up to now there exists a number of projects developed by ESA which plan to use radionuclide power or heat sources. One of the most far progressed is the 'Rosetta' project."
[Update: When this paper was written, it looks like they might have been only planning to use an RHU to heat the lander:
"'RoLand' includes not only the solar power generator but also RHU on plutonium-238 of the RHU 'Angel' type. By the opinion of the project experts, the RHU use is the most desirable, sensible and reliable means for achieving the main goal of the mission."]
[0] Energy supply for deep space missions: Risks of nuclear power in space and prospects for solar alternatives. From http://www.ianus.tu-darmstadt.de/media/ianus/pdfs/arbeitspap...
http://en.wikipedia.org/wiki/Radioisotope_heater_unit
never RTG (Radioisotope thermoelectric generator).
"While both RHUs and RTGs use the decay heat of a radioactive isotope (usually Pu-238), RHUs are generally much smaller as a result of omitting the thermocouples and heat sinks/radiators required to generate electricity from heat."
In aerospace engineering, they are used in just a few circumstances to augment the active heaters in a spacecraft: Heater lines, which are resistive heaters in a foil with a PI controller. You need active thermal control since your environment changes a lot (launch, preliminary orbit, transit, orbit) and thus its thermal parameters (some parts of a spacecraft may experience temperature gradients of a few hundred Kelvin within a few minuts). Most spacecraft designs have issues with overheating and thus RHUs cannot be used here.
An RTG's primary objective is to produce electrical energy (at an abysmal efficiency). The excess heat generated by an RTG is seldomly used for thermal control since it can also not be regulated and it emits simply too much heat energy.