The limits and potential of paleogenomic techniques for reconstructing grapevine domestication