A brief history.
At one time most hobby RC control was on 35MHz and 74MHz. The frequencies were allocated as to what you were using, i.e. 35MHz was generally used for flying things and 74MHz for wheelie and floaty things. On these frequencies there were set analog channels which meant that a limited number of people could fly, roll, or float together at anyone time in the same area.
As digital electronics and UHF RF devices got smaller, and more importantly cheaper, manufacturers started producing digital 2.4GHz equipment. These systems provided a high degree of immunity from co-located interference allowing for more users to do their thing. As these systems matured so did the modulation methods and protocols thus increasing the robustness.
Up until the introduction of the Phantom 2 Vision and Phantom FC40, 99.9% of RC control was on 2.4GHz, though the long range guys used 433MHz and 459MHz. Both the P2Vision and the FC40 used 5.8GHz for control as the fitted cameras used 2.4GHz WiFi to stream the video. The reason for this was that at this time there were very few, if any, mobile devices capable of using 5.8GHz WiFi.
Today it is generally the norm for WiFi controlled drones to be able to switch between 2.4GHz and 5.8GHz. The WiFi devices used are cheap and generally both frequencies are supported in the same package, also the manufacturer doesn’t have to re-invent the wheel as they can just fit prebuilt off the shelf WiFi devices into their drones.
Both frequency allocations have their own strengths and weaknesses. These are dictated both by the laws of physics and by the regions they are used in.
In Europe the CE standard is that 2.4GHz is limited to 100milliwatts to the antenna, and 5.8GHz is limited to 25milliwatts. Where as in the USA 2.4GHz is allowed 400milliwatts and 5.8GHz up to 600milliwatts, to the antenna. The 2.4GHz, power for power, offers greater range than 5.8GHz due to the greater absorption effect as you go higher in frequency. But the 2.4GHz band is only 80MHz wide and has many users. This means that the bandwidth available to each user is small. In realistic terms this means that the received live video will be lower resolution and have increased latency due to the lower data rate. On 5.8GHz the amount of spectrum available is in the 100’s of megahertz. This means that the amount of data a user can send is much greater, generally resulting in a higher resolution of the received image and less latency. However in Europe the range is significantly less than than that of 2.4GHz because of the piddling candle power limits allowed on this allocation.
Now I could wax lyrically about the inverse square law and how to calculate path loss, but only if you start to piss me off. 
Nidge.