Two separate aspects of orientation.
As for the entire mic in free space- When using the grids, the small size of these things makes them more omnidirectional than larger omnis, but there is a very slight off-axis roll off of the highest frequencies as you move from away from the front of the mic back around towards the cable entry point. Orientation of the internal rectangular diaphragm and holes in that case does not matter.
As for the orientation of the internal rectangular diaphragm with holes on one side- The only time I concern myself with the orientation of the capsule holes (meaning I take the grid off to look and see which way the holes are pointing, turn the mic if necessary and replace the grid) is when I'm either using the mic in the hard rubber boundary mounts or taping it to a soft surface. With the boundary mounts you need to make sure that the mic is inserted so that the holes do not face down, away from the tiny slot in the top surface of the rubber mount. DPA recommends inserting the mics in the rubber mounts so that the rectangular diaphragm is perpendicular to the slot opening. I insert the mic so the holes face up directly towards the opening because the boundary mounts increase the response peak slightly and it sounds slightly less peaky to me that way. It's been awhile since I used those mounts, but if I recall correctly, I think I preferred using them without either grid, just the nekkid capsule in the mount. I usually use the short grid for almost everything else.
A more common scenario is when I tape the mic to a soft surface like skin, fabric, foam, or whatever. I usually orient the capsule so that the holes would point away from the surface. Not because the mic is more sensitive on that side with the grid on, but because I don't want the surface to deform and obscure the side of the grid where the capsule holes are, changing the response (by acting like the slot in the boundary mounts). It's probably not an issue at all if taping the mics to a hard, non-deformable surface like a wall.