* For vertex shader output only, we do the streamout as a point list and
render each vertex once then we can reuse the same(ish) index buffer and
topology as the original draw.
* For GS/DS out there will likely be some expansion of verts so we do the
same as we used to.
* This means when multiple fragments are writing to a pixel you can choose
precisely the one you want to debug, rather than the debugging always
running the approximately last fragment to pass
* This means that e.g. decimal separator will always be . and similar
effects, which avoids the need to have culture specific formatting or
special-case handling around CSV export etc.
For now this just acts like a really wide vector with N columns for the
array, but this could be improved in future if a better visualisation is
decided on.
baldurk