in most cases where ABS kicks in, a non-ABS driver will skid – losing control and increasing the distance they travel

That’s the “lowest-common-demoninator” argument. I’ve made hundreds of panic-stops over the years, and never have I lost control of the vehicle. And I don’t really consider myself to be a “great driver” (certainly my accident history in my younger years suggested otherwise).

Also, the notion that losing control and skidding will increase the stopping distance is, according to the NHTSA anyway, incorrect:

A vehicle yawing out of control with its wheels locked may stop in a very short distance, while a stable vehicle (its directional control maintained throughout the duration of the stop) may require a very long distance to complete its stop.

That said, the NHTSA document cited above has a lot of interesting information on ABS performance, and the discrepancy between test-track and real-life results.

Numerous crash data statistical analyses conducted over the past few years… indicate the apparent increase in single-vehicle crashes involving passenger cars equipped with four-wheel ABS almost completely offsets the safety advantage such vehicles have over their conventionally-braked counterparts.

Another interesting note:

This study also establishes that antilock brake systems include compromises of stopping distance versus vehiclar stability. Most antilock brake systems maintain vehicular stability while braking by minimizing excessive yaw. In a curve, this stability may be created by sacrificing the shortest attainable stopping distance. With this said, most test vehicles… stopped in shorter distances with ABS… for maneuvers that involved braking and steering…

And finally:

We believe this study has established that ABS braking performance deficiencies are not responsible for the apparent increase in ABS-equipped, single-vehicle, run-off-the-road crashes.

So… they say it works fine on the test track, and still don’t know why people are more likely to die in ABS-equipped cars than in non-ABS cars. Finding out what’s up is still on NHTSA’s to-do list:

It had been predicted that ABS could prevent numerous crashes, assuming drivers knew how to use the systems. However, statistical analyses of real-world collision databases suggest that the introduction of ABS does not reduce the number of automobile crashes where it was thought ABS would have proved most effective. Crash studies show increased involvement of ABS-equipped vehicles in single-vehicle crashes and less involvement in multi-vehicle crashes. Specifically the increase has been in single-vehicle run-off-road crashes such as rollovers or impacts with fixed objects.

Anyway, based on the above I’ll concede the point that stopping distances don’t necessarily suffer with modern ABS systems (the ones tested in that report). But ABS is still a net-negative in terms of safety.