It doesn't really need it and it would needlessly add to its mass. Rosetta will assume a relatively slow and probably highly elliptical orbit around 67P/Churyumov-Gera...aaagh! Chury!, with the perigee at only roughly a kilometer away from it. The orbit has not yet been determined though, see my related question and the answer there. It likely won't be until Rosetta transmits its own close proximity observations of the comet, and might even change in time as the comet's coma, tails and in general its surface activity increases during its flyby closer to the Sun. Rosetta of course also has its own semi-autonomous collision avoidance system onboard, that will enable it to react to any debris in its path and adjust its trajectory.
Anyway, such elliptical orbits leave mission control plenty of leeway to later (when the Rosetta will be in orbit) decide at what point to release Philae on its ballistic trajectory. My guess would be this trajectory will be attempted from the point of Rosetta's flying past the comet in the opposite direction of its movement, to reduce the chance of incoming debris to near zero. The lander also has a harpoon (see the spike in the middle of the legs frame) that will fire towards the comet immediately after the touchdown to hook itself on it. Additionally, each of the three legs have battery powered ice screws in between the two padded gears, to additionally grip to the comet and pads to cushion the collision / landing:
Training sessions for the Philae comet lander (Source: DLR, National Aeronautics and Space Research Centre of Germany)
To answer your questions more directly tho, adding propulsion system to it would be needless weight. The Philae lander wouldn't be able to react fast enough with its Attitude Control System (ACS), if it had any onboard, and to avoid any possible obstacles. Instead, the lander will be put on a collision course (or ballistic trajectory, your pick) with the comet with an approximate velocity of 1 m/s. That is, I would say, higher relative velocity than lowest achievable with its release system, and the resulting kinetic potential should suffice to negate effects of any collisions with smaller debris or pressure in the opposite vector to its movement from outgassing, as you mention. The lander's mass is negligible in gravitational sense, but it still adds to its inertial mass. With a bit of smart decision making from the mission control, they ought to have plenty of opportunities to assure its touchdown. The harpoon, soft pads and powered drills should do the rest in assuring the lander stays put where it landed.
Of course, there are chances all of this will go sour. One of the main concerns I've seen mentioned online is, that the lander itself creates with the harpoon and the screws deep enough cracks for a part of the comet on which it landed to simply chip away from the comet's main body. That doesn't sound too serious as the lander could still do all of its experiments (some might become less conclusive tho, like for example the CONSERT comet nucleus radio wave sounding experiment), just on a smaller piece of the icy rock, but it could cause the lander to completely unhook from any piece of the comet's surface, or the chipped away piece starts spinning, squashes the lander between the chipped away piece and the comet's main body and causes it to lose its grip to it.
I guess we'll have to wait and see, but adding propulsion system to it wouldn't have mitigated main concerns I've seen mentioned over its chance of success either. In fact, it would only add to the problem, with more parts that could malfunction and larger mass of the lander increasing its physical influence on the comet.
