Gates is right.
Almost all climate occurrences are thermally activated.
This argument requires a grounding in statistical physics but here goes.
Assume that some physical process is activated thermally and it is described by an activation energy governed by Boltzmann statistics.
If the activation energy is 0.46 eV, which is around the value for water vaporization, then you can get changes of probability of 15% for a 2 degree change in temperature.
http://www.wolframalpha.com/input/?i=exp%28-0.46%2F8.617e-5%2F275%29%2Fexp%28-0.46%2F8.617e-5%2F273%29
If the activation energy is greater than this the probability changes are greater due to the steepness of the barrier. Higher activation energies are associated with stronger bonds.
Now consider that probabilities are multiplicative and if we have events that require lots of mechanisms of reduced probability, that 15% can start making an impact. And that is all from a 2 degree change, with enough time for the change to have an impact. This has positive feedback aspects should the warm seasons start to become longer as well.
Anyone that has done industrial or laboratory furnace control understands how important it is to maintain a stable temperature. This is something to really think about.