According to the article, in early days, wind generators were simply tripped off line if there was a voltage transient, because the power companies didn't want to deal with their unconventional characteristics and there was so little wind capacity on the grid anyway. Then, as the fraction of wind power grew (in Europe, first, and then recently in New Mexico where the grid can get 20% of its power from wind under some conditions) it because necessary to actually solve the problem of making the grid work reliably with a mix of conventional and alternative energy sources. Finally, the U.S. Federal Energy Regulatory Commission adopted the German E.ON Netz Fault Ride-Through (FRT) specification.

It turned into a bit of an argument, because the FRT spec was tough for the wind generators to meet, and their manufacturers thought they were being discriminated against, while the conventional generator operators thought that the windmills were being given a break. The crux of the argument is that the specification required the generating plant to ride through 0.625 second faults where the line voltage has dropped to 15% of nominal. What it doesn't do is require ride-through of any amount of zero line voltage.

The conventional machines want a load, and when the line voltage drops, their load drops and the machinery speeds up. The system can ride through a very short period of zero line voltage, but if it drops to 15% for more than a second or so, the generator shaft gets too far ahead, the system loses synchronization, and the protective circuits trip it from the grid.

The windmills have variable speed shafts (depending on wind conditions), and the AC characteristics are controlled by electronics. So while it's expensive to add the capability to ride through a drop to 15% voltage, when you've done it, you can ride through much longer failures.

So now the problem boils down to the fact that the new windmill systems have the ability to handle the new FRT spec, but the older conventional systems don't--and it's really hard to retrofit them. All sorts of mechanical systems like lubrication, feedwater pumps, cooling systems, fuel compressors, active seals, etc. are all very sensitive to power fluctuations and have protective circuits that trip if there's a line fault. It's not that the technology is unavailable, it's just that it's not factored into the cost of the plants because it wasn't included in the original system.

It's kind of a weird problem: You want to improve the reliability of the grid while adding new power sources, but when you do so it exposes shortcomings in the existing physical plant. This makes it tough to calculate the relative cost of the systems, because even if you eliminated wind from the equation altogether, you're still stuck with these old plants that suffer nuisance trips all the time.

Note: I don't actually know anything about this, I'm just summarizing the article...  :-)