Player X switched in Scizor!
Player Y's Mamoswine used Ice Shard! (it does crap)
Player Y switched in Shanderaa!
Player X's Scizor used Bullet Punch! (it does crap)
Player Y switched in Doryuuzu!
Player X's Scizor used Bullet Punch! (it does crap)
Player Y's Doryuuzu used Swords Dance!
Granted, in this particular scenario, Magnezone would work, too, but you get the point. Shadow Tag is basically a suspect for the Support characteristic, no matter how you look at it.
Sure, you can concoct theoretical scenarios in which Shandera sets up and sweeps, helps a teammate set up and sweep, revenge kills two or three foes, etc., etc., but they carry about as much weight as the bad shandera play in the youtube video posted (less, even, since at least that was a real battle). I could just as easily concoct scenarios in which the foe mops the floor with Shandera all day.
To make any meaningful statement about shandera's potential in the metagame, there needs to be a metagame. You can't consider just a few pokemon and think you've got shandera pegged (or even a definite suspect for being pegged). The metagame is a complex system and you can't figure out part of that system by cherry-picking a subset of it for your analysis.
For all you know the metagame will evolve such that choice band scizor rarely gets used (maybe because it's set up bait for ST shandera, or maybe for other reasons). In that case, your scenario doesn't hold much water. Yeah, it could happen, but unless it or very similar scenarios happen much more frequently than 1 in ten battles or so, it really isn't all that significant.
You could very well be right that ST shandera ends up a candidate for suspect testing, but it could just as well end up reasonably balanced. Only time, and testing, will tell. In the meantime, all this theorymoning is just a waste of time and breath (or some digital, non-spoken equivalent).
...especially if ST Shandera never gets released.