Titrating a mixture of strong and weak acids will yield endpoints such that the volume of base required to reach each endpoint will be different for each acid, e.g. in a mixture of HCl and CH3COOH, the volume of base required to reach the HCl endpoint will be less than that for the CH3COOH endpoint. This is all down to relative acid strengths.
However, the above isn't observed when titrating a polyprotic acid, as each endpoint requires the same volume of base; this is counter-intuitive when considering the above paragraph. For example, H2SO4 also has two endpoints – one for H2SO4 and one for HSO4- – but the volume of base required to reach them remains constant. This doesn't sit right with me, considering that HSO4- is a weaker acid than H2SO4.
So my question is: why does the volume of base required to reach each endpoint in a polyprotic acid titration remain constant, even though each endpoint will result in a weaker acid? Surely the volume of base will increase each time, like is the case in a mixture of stong and weak acids?