Would require another spin of the board, but the thru-hole size seems, sadly, just TOO small for a standard RPI header to pass through. Would request/recommend that this be fixed on the next iteration, if there is one. (are your schematics, etc, open source?)