It depends on the shanpon wait imo. The potential for suji trap won't be enough against more experienced players who have the discipline to play full defense. I'd probably just choose kanchan wait regardless, but that's not a very strong play in the second example hand. The only time I would choose shanpon is if one or both pairs were kabe-blocked by other tiles (such as waiting on 1p/4s when all four 3p have been discarded).
I saw a mahjong pro (Tsuchida Koushou) who has a "toitsu theory" translated on osamuko, an english mahjong blog site. That theory, very roughly, is as follows:
Most likely pairs are 1, 9, and off winds
next are 2, 8, seat wind
next are 4, 6, dragons
next are 5, round wind
least likely are 3, 7
There are additional things about hand/tile shapes and the idea itself here (very interesting even if you do not agree with it!):
http://www.osamuko.com/tsuchida-koushou-homage/
The only reason this theory, which sounds like total occult bs, is talked about at all is because tsuchida koushou was one of the strongest professionals in Japan, and because it's really just a way of organizing information. Middle tiles and the 3/7's are more likely to be useful for player's hands, and honor tiles are more likely to be kept until later in the game, meaning you have less information about whether they're still in the wall or not. Edge tiles are more likely to come out first, meaning that if they still haven't been discarded later on in the game, it's more likely that they're still in the wall or that they're being kept to go after a yaku like chanta. Even though this is a chiitoitsu theory, it can apply in much the same way for getting extra tiles of the same group through understanding where the other tiles are (used by other players vs. in the wall).
Basically, in both examples, even with the chance of getting san an kou, i'd still take the kanchan wait almost every time. It has equal potential of naturally creating a suji trap, it doesn't assume that you'll draw a tile that will improve your wait, and it's still a hand that has a good chance to get mangan in both cases. If you wanted to distill it into a few general guidelines...
-1/9/off wind shanpon is ok
-kabe-blocked shanpon is ok
-suji trap shanpon is ok against opponents who rely too heavily on suji
-otherwise, kanchan is less likely to leave you in karaten (empty tenpai)