Stag,
Water, like silica glass, is highly opaque to thermal sensing -- both substances limit / block thermal radiation (in the wavelengths detectable by most thermal sensors) from passing through. Water will absorb, while silica glass will reflect the thermal radiation. So, yes, if condensation forms on the objective lens of a thermal scope, it will degrade the image.
If you use a "clip-on" thermal imager, you need only to establish "zero" on (i.e. boresight) the day scope that you are using with the clip-on thermal imager. The optical collimation on the thermal imager will allow the boresighting on your day scope's aiming reticle to be used.
-----
Enclosed below is a comparison of images of the same geography taken with different imaging technologies - thermal sensor at the top, night vision (Gen 3 Pinnacle light intensifier) at the bottom, and a digital camera (unaided) in the middle. These images were taken within seconds of each other on December 31 at around 11:00 PM, on an evening with clear skies but no moon, and which appeared very dark to unaided eyes.
The thermal imager (clip-on) is one that I had constructed on an ultra-high sensitivity, high-resolution, thermal core and which I have previously shared images from at this thread:
http://www.snipershide.com/shooting/snipers-hide-night-vision-devices/215301-thermal-vs-nv.html
The conditions (low temperature and low relative humidity) were close to ideal for thermal imaging. In the thermal image, note the detection of the power lines on the tree-line at the horizon, which is over 800 yards away. The power lines are just over 1 inch in diameter each, and are not visible during the daytime at this distance (or even 1/4 the distance), against the heavily forested background, with the naked eye, even under direct, bright sunlight.
The i^2 night vision image was seriously degraded by glare and blooming from street lights even though they were between 500 and 800 yards away from the view point.
The night vision scope, a PVS-14, and the camera lens used for the "unaided" photo, both have a much broader field of view (40-degrees), and therefore I joined two images captured from the thermal scope for the comparison.
IR-V
Compact image (thermal, unaided, night vision):
Large image (thermal only):