It takes longer for a BJT device to come out of saturation than a MOS device to come out of triode region. It is explained by the base charge storage time, the charge has to recombine and go back to emitter and collector at the same time. Isn't this the case for MOS devices too? The charge in the channel has to go back to source and drain. What is the cause of BJTs being slower to come out of saturation than MOSFETs?