I could have sworn that I recently read posts here and elsewhere, and articles in various places, that emphasised the value of human pilots and other aircrew on current operations. How else do you explain the fact that ISTAR types acquired in recent years?
LOS with a manned aircraft flying 3-500nm away does not give latency issues (not that 1.5 secs (max) with Geo Stationary Satellite is that big an issue anyway for most applications!). Furthermore, the latest micro-satellites in NEO from UK manufacturers could be provide a better solution or the 60,000ft balloons also already pointed out.
Satellites cost money to build and launch. Ones in Low Earth Orbits have a restricted field of view/footprint so you need more. The Iridium system has 66 of them - not including six spares. Balloons may be cheaper to launch but how many do you think you would need? This sort of infrastucture won't be cheap. I almost forget to mention the effects of bad weather which tends to attenuate microwave signals. What happens if the one time you really need the system the monsoon has reduced the data capacity (look up Shannon-Hartley law) to less than that you need?
But those 'pilots' are not cheaper. Meatware in an aircraft with accommodations for same is more expensive than remotely piloted air vehicles, unless you can economize on the little things such as training, pay, pensions, and search and rescue operations.
Surely a ground based pilot still needs training, pay, pension, etc? As for SAR, surely you're not suggesting that aircrew are the only service personnel who may need rescuing or extracting from somewhere nasty?