That's an interesting link, thank you.
But taking your example, at 2,000 feet the signal from a cellular antenna, most likely with a downtilt at almost ground level a mile away, should be extremely small. It shouldn't be too difficult to engineer the cell sites accordingly.