... except that even at that speed, with the timeframe (roughly) known and with the sightlines (roughly) known, Clayton was NOT out of visible range. If we estimate 4 seconds before impact to be "decision time" for the cop to make the U-turn or not, he would have been about 160 metres away. You can easily see a headlight at night that far away. For that matter, even if the cop had seen the headlight, 160 metres and closing (even if the cop didn't establish the closing speed), would have been close enough that, had the cop seen the headlight (and either misjudged or didn't bother establishing a closing speed), the cop ought to have recognized that his U-turn would take longer than the approaching vehicle would take to be right on top of him.
Actually, we do not know the timeline at all. They could estimate his speed at braking and wheel lock up, but not before. How do you know he wasn't doing over 200 kph at the beginning of the sightline? Not to mention the fact that Clayton would have been behind him, so that further aggravates visibility as does the construction area.
The cop could have very easily established that there was an approaching vehicle and that it would be best to defer the U-turn until after it had passed.
The most likely explanation is that the cop didn't bother to look, at all, prior to making that U-turn - or if he did look, it was nowhere near an appropriate timeframe before making that U-turn.
Apparently the investigation did not reveal that. The point being argued is in the end of this sentence.....appropriate time frame before making the u-turn. People keep arguing if the bike could be viewed from that distance, but even that has some context being deliberately ignored by some. To stop and look directly up the road, sure visibility is wonderful. Unfortunately, the sight over distance accuracy is directly related to both the real-time circumstances in which it was viewed, as well as the real-time rate of changes to the vehicles being viewed (speed, where on the road they are). It is not remotely uncommon to not see a vehicle while doing a shoulder check, which is why that defensive manouver is pretty much useless without being accompanied by a second shoulder check, all while monitoring mirrors to have a sense of what is approaching you, and this is all to deal with people moving in the same direction and relative speed as yourself.
happycrappy, your example is a poor one. Being rear-ended is the rear-ending driver's fault. The Vienna-convention traffic rules are spelled out. The presence of a large, slow vehicle in the midst of a maneuver doesn't give another driver who has priority the right to hit it ... but it also doesn't give the driver of that large, slow vehicle carte-blanche to pull out in front of another motorist who has priority, and certainly not to pull out in front of that priority vehicle in such a manner that they cannot stop in time. In a situation as described here, both drivers would be held at fault - the priority vehicle (bike, in this case) for speeding, the U-turning vehicle (cop car) for failure to give way ("yield") to the priority vehicle.