I have no experience from these kind of cameras in real life. But in the article you reference it says:
"it is a good thing that the thermal imager can see through fog to a certain extent."
Sounds to me it is not totally unaffected. Also, the thickest fog in arma is quite thick and would deliver a massive amount of moist in the air. Seems reasonable to me it has a severe impact on the way heat can be read at a distance.

EDIT:
Found this folder from FLIR which backs this up as correct behaviour:
http://www.flir.com/uploadedFiles/FOG_techNote_LR.pdf

"Humid air acts as a “shield” for infrared
radiation"

Here is another link, suggesting thermal imaging outperforms visible wavelength in light fog, but is no better in heavy fog.
http://www.securitysa.com/article.aspx?pklarticleid=5369

Currently in Arma3, thermal is no better than visible wavelengths in any fog conditions, which does not match reality described in these links.

It depends on the thermal imager, different wavelengths of light are attenuated by different things. But since ARMA 3 takes places in the future I would assume the NATO thermal cameras would be two color or more. This would mean they would be less effected by atmospheric conditions if the wavelengths they detect were chosen carefully by the engineers.

Mass closing ancient tickets with no activity for > 12 months; assume fixed or too trivial.

If this issue is still relevant in current dev build, please re-post.