There were some incidents of interest in Austin.
1. The Pérez–Sutil accident on the opening lap
Sutil's Sauber was parked on the kerb, and the track was covered with debris. Race control immediately deployed the safety car – a good call.
Hopefully, even with a virtual safety car system in place, the real safety car will be used in incidents like this when track workers have to remove a lot of debris. Bunching up the field makes sweeping a lot safer.
2. Speeding behind the safety car
A number of drivers were penalized for speeding when the safety car was on track, and they were only given a five-second penalty. I think the five-second penalty is excellent for small offences, but if the FIA are serious about safety, they'd better start handing out drive-through penalties for speeding behind the safety car.
3. The recovery of Hülkenberg's car
The car was abandoned along the armco some 300 metres down the straight between T11 and T12. It was completely outside the track and close to an exit in the barrier, but the runoff is very narrow.
It took quite some time (two or three laps) before the marshals got to the car. Normally at a grand prix, you would see them rush out as soon as the car has stopped. I suspect they have been instructed to be more cautious. That's good.
What I found disturbing, however, was that the cars were going by at high speed. I assume there were double waved yellows, yet they didn't seem to slow down much at all. Charlie Whiting has already said there has to be a better way of making the drivers slow down under yellow flags. Two grands prix after the tragedy at Suzuka, nothing has changed.
I realize they can't introduce a virtual safety car system until it has been thoroughly tested. But why aren't the drivers told to slow down more? There are already rules for how much slower they must go under yellows. Just increase that delta for the remainder of this season.
And please consider the introduction of local slow zones with a speed limit for next year. Force the cars to go by a recovery scene at a safe speed, say 100 km/h. At a high-speed part of the track where cars average 310 km/h, you would only get them down to 230 km/h with a 35-percent increased delta time. That's not a safe speed for either drivers or marshals – not even in the dry with tarmac runoffs.
Regarding (1), it wouldn't make that much difference on the opening lap since the cars were still in a fairly tight pack anyway. But yes, I agree that sweeping the track should normally only be done when there's no race traffic coming, with the post chief calling his men off the track before the cars come by.
Regarding the bolded part, I'm not satisfied the FIA has got to the bottom of this network of problems yet, and I think they know that, which is why they're very sensibly waiting for the accident report before changing the rules. We have to remember that respect for yellow flags, in terms of slowing down, has been very poor in F1 for a long time, and the wrong signal has been given to drivers for a long time (like Interlagos 2003 when Alonso was awarded a podium finish which he would not otherwise have got after he ignored yellows, and caused a huge, race-ending crash under yellows that endangered the lives of Mark Webber and several marshals at the scene. Not only did that go unpenalised; Alonso went home with a better result than if he had observed the flags and avoided the accident.
Yet despite this appalling situation having persisted for many years, serious incidents under local yellows have been very rare, so we shouldn't exaggerate the risk in just leaving things more or less as they are for a couple of races until the panel has had a chance to report.
Maybe Charlie has indeed insisted on marshals obtaining his permission before going trackside to recover cars, but if so, I'm not as convinced as you seem to be that he's made the right call. Sutil's car was ideally placed to be pushed back behind the barrier, but Sutil was reluctant to leave it because it was in neutral and it was rolling back down the track. If one marshal had gone to help him steer, they could have got the recovery done in a few seconds without machinery. As it was, the car rolled to a point where it would have had to be pushed forwards, back up the hill, before it could be steered back into the gap, and as the television pictures cut away they were bringing a digger in (I'm not sure if they used it in the end - there was clearly no need). Whether they were following instuctions, I don't know, but when you've got a car rolling back down the track past the access gap, a driver standing by the car not being directed to a place of safety, and three marshals standing behind the barrier doing nothing, that to me looks like it's simply extending the amount of time the car and driver and (eventually) marshals are going to be exposed to danger. The situation was nothing like the Bianchi one, it wasn't necessary to have a SC or a slow zone, because the chances of anybody going off under local yellows there, in those conditions, were just vanishingly small.
That's why we need to wait for the report and hope it looks at risk management as a whole, rather than just focussing on the thing that's just happened. If we start saying marshals can't assess the situation for themselves in a case like that and go and get the car before it rolls past them and becomes harder to recover, then we've created a rule that mitigates a risk that's so small it's virtually non-existant while still allowing, for example, flag marshals to stand on the pit wall during the start. And if somebody stalls, the marshal standing just feet away from him sticks his head arm and shoulder trackside of the debris fence to wave his flag, even though if the stalled car does cause a collision, they're at serious risk of getting a wheel in the face. Recovering a car from the outside of a fast bend in the wet when several cars are on heavily worn inters it would be different, but on the inside of the track in the dry, come on. Just get the track cleared asap and get on with the race. The race director needs to trust the marshals to know when it's safe to enter a live race track and when it isn't. Maybe that is still the case and the marshals at Austin were just not used to being expected to go on a live track. Maybe they normally work ovals, or maybe they're just more familiar with the system used by IMSA and Indycar and USC and every other North American series where you just call the SC for everything.
Whatever, we need to try to consider as many risks as possible and rank them for likelihood, severity and how easy they are to mitigate without impacting too much on the racing, and address the biggest risks, not just the thing that happens to have occurred recently.