The X Y configuration is more accurate useful when you want to slightly de-emphasise what is in the centre (as cardioid capsule of the stereo field, which is facing forward) and great if you can vary are recording content to work with dialogue.
There is only a subtle centre de-emphasis but it is enough to help with improving the stereo width perceived clarity of the overlaid dialogue, if the dialogue is in post productionthe centre channel or panned centre.
X Y is generally easier
The M/S gives a lovely centre clarity to implement as no "decoding' (sum stereo recordings and difference) is required, although ideal when you really want to emphasise the centre of a stereo recording and you can still automate this in post so that you can vary the width by changing the angles to make them a near coincident pair.
Cost wise figure amount of eight mics tend to be more expensive than cardioids if you want them perceived direct to be smallindirect sound.