Midwest, yes, according to my reasoning that is my expectation. I think it is the electrical disturbance/transients caused by the action of switching the cooker hood control (whether up in speed or down) after the motor is run up that precipitates the RCD trip via the short leg of the ring final circuit. It may be though that the transients are of different strengths and forms depending on whether one steps up or down in speed. The only real test is to do it and to to connect up test equipment to analyse why.
All that said, I am conscious that these types of problem are nightmares for the busy electrician and wasteful in time and money. That is why I tried to offer a practical solution first - the filtered socket and maybe transient tolerant RCBO. Be interesting to know if one or other solved the problem.
Rob190's idea of connecting the hood 13A socket further around the ring ( - in the ring or as a spur) was the simplest and cheapest and thus best way to proceed. It could be easily trialled first too.