1. You would have to know the hematocrit to correct for variation and I don't think there are any methods that use the same dried blood spot (you would have to collected liquid blood and that defeats the purpose).
2. From a newborn screening perspective, a full circle (the marked dashes) is approximately 75uL. If the blood is well within that zone or outside of that zone, then you are approaching 100 or 50 uL respectively. But the blood volume from an actual punch shouldn't vary substantially within these volumes and close to filling the marked circle. There are manuscripts from the CDC that have been published that describe these variations - check out the CDC website - search for newborn screening or look for papers by B Adams and W H Hannon in ClinChem which will be most helpful to you.
One solution, if not newborn screening, is to use a calibrated pipette after collecting in a syringe. If you are doing this you might consider adding internal standard (stable isotope labeled if using MS) to the blood before spotting.
3. Internal standard for DBS varies by metabolites you are measuring. As mentioned above, if using MS then I suggest a stable isotope enriched analogues of your metabolites. If not then use standards typically used in your assay.