I have done some work on near-IR (1100-800nm) using a Fujifilm Finepix IS Pro and I can confirm that the performance in haze etc is markedly better than visible light.
As far as longer wavelengths go, whilst the atmosphere does attenuate signals a lot, bear in mind that big, close, hot targets like the sun start out with an enormous intensity per square whatever above the atmosphere; even attenuation by (say) 60-100dB still leaves you with a lot to play with.
I am reminded about a local AM radio transmitter that had an ERP (power) of something like 500,000W. Local radio hams were having problems with the second and higher harmonics, which fell into ham frequencies. The site operators took the concerns seriously, checked everything out and confirmed that their filtering was actually rather better than required by law. However, they were still radiating a goodly fraction of a watt on the amateur bands - 60dB down on 500kW is still half a watt - hence the locals having difficulties. It was never resolved; there wasn't much the site operators could reasonably do and they were already cleaner than they needed to be. It's the same with the sun: even a lot of attenuation leaves a detectable amount.