His conclusions seemed suspect to me - I remember doing out the calculations myself and came out with the conclusion that a 50x50 mile solar array in the Arizona desert could power the U.S. A 150x200 mile array in the Sahara could power the world. These are large arrays - they would be visible from space - but they are far from "country-sized", and quite doable if production costs come down (which they are - solar and wind are now cheaper than coal).
(You can check my math yourself: total world energy consumption = 109,136 TWh [1] / 8760 hours/year = 12.45 TW / 15 W/foot solar panel output [2] = 830B square feet / 27878400 ft^2/mi^2 = 29772 mi^2 = 150x200 mi array.)
Working through his math, I think the problem is that the UK is unique in a.) having a high population density b.) with a large per capita energy consumption and c.) a lot of cloud cover. On his chart, only Japan, South Korea, the Netherlands, and Bahrain had it worse. The U.S. and China both had an order-of-magnitude better energy consumption per unit area; most of the world had two orders of magnitude. That's an argument for why Britain is fucked in the new globally-warmed, fossil-fuel-limited world, but not why the world is fucked. A lot of developing areas could see their energy consumption rise to rival the U.S. and still not bat an eye with solar power.
The EROEI of solar panels is between 8.7 - 34.2; wind is 20-50. [3] As long as this number is > 1 (which it is for nearly every energy source except biofuels) you can keep building production capacity indefinitely; you just invest the energy from previous capacity into building new capacity. Higher numbers of course let you do more other stuff while this is going on and build out faster, but it's just like capitalism. As long as you turn a profit on each panel, you'll have more as time goes on.
[1] https://en.wikipedia.org/wiki/World_energy_consumption
[2] https://www.solarpowerrocks.com/solar-basics/how-much-electr...
[3] https://en.wikipedia.org/wiki/Energy_returned_on_energy_inve...