Well, the article makes for very interesting rhetoric: poisoning the well for people arguing against telework by equating it with categorism. The heavily-loaded language is a dead giveaway.
But without data to back it up, it's still just rhetoric. The fact is, even a relatively small group of people can ruin something for everyone. Ask anyone who has worked on teams for any length of time, and they'll tell you just how much damage even one person can do.
How do you predict who these people will be, or what kind of damage they'll do? By and large, you don't. That doesn't mean people don't try: the so-called "poisonous people" phenomenon has plagued the open-source community for long enough that entire books have been written on the subject. But the problem remains, and there's still no fair and reliable way to predict who will turn toxic.
Without that, you've basically got two options. You can decide that the costs associated with abuse are worth preserving the privilege, and mitigate the damage as best you can. Or, if the damage is too severe or the problem too persistent, you can declare that the poisonous people have ruined it for everybody. Yahoo took the former approach for a long time, and it hasn't worked out for them. Now they're trying the latter. Is it truly fair to blame them for that?