Re: Detection would be a good start
Firstly, most modern systems don't use triangulation, but time difference of arrival, as I think I mentioned.
I agree that you may need quite a few stations to track signals on the ground depending on how large an area you want to cover, but for tracking the actual drones the problem is a lot easier since we are only interested in the skies on the approach and departure flight path. Clearly, we can construct antennas that point in that direction and ignore most of the interfering signals that don't interest us.
I am not convinced by your interference argument. Consider the drone. It is clearly flying at a reasonable altitude in order to strike an approaching aircraft. If it can distinguish its own control transmitter from the hundreds of other 2.4GHz sources visible from the air then so, presumably, can our tracking system.
I don't believe that there are any mobile phone frequency bands around 2.4GHz and your complaint seems to be based on the rather quaint idea that mobile phone locations aren't already tracked. As for Bluetooth, this does sound like it would be too low power to be usefully tracked.
So, in summary I don't agree that this is not a bright idea.
In related news, a colleague has pointed out to me the following system which seems to take the special to purpose radar approach: http://www.blighter.com/products/blighter-auds-anti-uav-defence-system.html