I think it'll be a pretty tricky thing to calculate, as the possible speed of motion may well depend largely on the viscosity of the molten rock down there where the plates are doing their thing. My bet would be that even a fifty degree C change in temperature could make a huge difference in viscosity up near the melting point of the rocks in question. It certainly won't be a linear relation.
Even calculating the frictional heat generated by plate motion would be equally tricky, as you don't have a good number for the coefficient of friction as a function of temperature or viscosity. But don't give him all 6000 years for his plate motion to take place - the people in India would have noticed all those earthquakes back 4000 years ago, when they were already keeping written records.