So like the active layer on top of the permafrost, it is the cycle of time from cooling winter to warming summer that makes the difference. There is too much mass of water to lose enough heat to freeze the arctic waters solid during the winter before the cycle of heating begins and melting goes on to summer.
During the ice age, the warming period is far smaller and the cooling is deeper and longer. Perhaps I'm too focused on the Arctic but this seems to prevent the release of more methane from hydrates. We found on the north slope that molecular structure is rather stable, trapping the methane and taking significant energy input to collapse the lattice.
Think Global, and don't ignore the pressure factor. As the seas drop because the water is tied up in miles thick ice sheets on the continents, the pressure comes off the hydrates (shorter water column above them). The hydrates in the continental margins get above the gas/water curve (especially in the tropics where temps are warmer), and de-gas, raising methane levels in the atmosphere. Global warming, in spite of high albedo from the ice sheets nearer the poles.
During the thaw, sea levels rise again, hydrostatic pressures on the continental margins are high enough to sustain hydrate formation again, more CH
4 goes into hydrates, and the process slows and finally peaks out, ending the warming cycle.
Of course, other factors such as volcanic activity, cosmic impacts, plate movement, ocean circulation, and especially solar output would affect the cycle, too, perhaps even more than the methane.
