The IPCC makes its estimates by 1) setting scenarios and 2) evaluating predictions of a range of models on these scenarios. This results in a 'likely' range on its fossil fuel intensive scenario between 2.4 and 6.4 degrees celsius, with a best estimate on that scenario of 4 degrees celsius.

So this is the upper bound of the likely range for the scenario with the highest emissions. In other words, Lovelock is assuming the worst case rather than the most likely.

That's his right, of course. But it would be silly to start taking it as the most likely picture of what's going to happen rather than a worst case possibility.