This may be trying to state the bleedin obvious, but the crash site does appear to be a selected site. I am no pilot but I am interested in human factors.
The reasons, as it appears to me, are
1. The surface was flat.
2. The area was large enough to put a helicopter down on (Even though the area was very tight)
3. The helicopter crashed away from the edges of the building (as dropping off or hitting the edge would be most likely unsurvivable)
4. The helicopter did not hit anything else on its descent
5. It was not populated by anything. Not a car park where there might have been cars parked awkwardly for an emergency landing to take place. Lamp standards may well make a car park look like a very difficult place to set down in an emergency without hitting one of them. This might not be so on a flat roof.
6. The helicopter landed approximately upright, (not inverted or sideways, spinning or any other position which would obviously indicate that the aircraft was well beyond any control)
So if I am correct, and IF is purely conjecture, then we might consider that the pilot had time to make decisions. This does not appear to be a situation where the pilot flew into the ground due to inattention or anything like it. The workload on the pilot must have been huge, but for each of the factors above, there may have been a conscious decision making process by the pilot. The maximum time for the pilot to make decisions would be largely outlined by the last radio message and the impact. That gives the outline, and the time between the aircraft encountering difficulties and the impact may well be far smaller, but at least there we would have some measurable time.