I have had the pleasure of realising that a crew were about to take off on a runway that was 4500' shorter than they were expecting. If you ever get out out of a first class seat and try to run to the cockpit door, you will find that it gets a bit of a response from the cabin crew, who knew I was an off duty PIC on type. I stopped before I got to the door, and then sat down on a spare F/A seat and strapped in again. Later in the cruise, I confirmed that the NOTAM existed on the ATIS, the NOTAM list, and I did a check of the performance that had been applied. I had done a quick maths while heading to bash on the door, and had concluded that they would get airborne just before the DER, and the perf data runs showed that had been the case. I was able to spot a landmark that was identifiable with the liftoff point. We were about 5' at the end of the concrete in this case, and the equipment was shorter than our gear. Simple case, a two engine aircraft will have a better margin to deal with a gross performance error than a 4 engine case, assuming that you don't lose an engine in either case. The probability of losing an engine in this case was probably around 20M:1. But that is one bird away from a bad day.
Today, my iPAD gives the notams for the airports, and highlights those in a list. They are readily identifiable as a taxiway, runway, comms, airspace etc matter, so the screening is far faster than the 88 pages of NOTAMS we got for a polar flight a few years back.
The event I was too close to arose where the ATC gave 5 different runways to the flight crew, and they were not a team that was known to tell the ATC to make up their mind. As an AAI for that company, I got to make my views to ATC later in the company investigation that occurred.
The NOTAM system is grossly deficient. There is no safety in a system that loses critical signal in the noise of the volume of inane routine messages. 88 pages, 15-18 NOTAMS per page, and the time provided by the company to review those is around 10 minutes at flight planning.