Sharing what I have learned from people who seem to know what they are talking about.
Total lock out - not likely with hopping protocols. When some of the 2.4 systems were on one or two channels and Spektrum was on DSM2, there were lock out reports that might have been related to a busy RF band, but it is hard to be sure of the cause without someone doing RF correlation to time in the air for the model. Apparently DSM2 was great when we had 200 pilot events and half were on 72 MHz but, as Spektrum said, as the events got bigger, they saw the need for DSMX. Now Spektrum hops too.
What about Increased latency? - This seems to be the most likely problem. I remember seeing reports of this when the first hopping protocols came out. As the band got busy the response time went up. I have seen this demonstrated on the RCReviews videos. And I seem to recall seeing reports of this at past big events. But it is hard to pin down the cause without someone doing a correlation between the model in the air and the state of the RF band.
Decreased range - General opinion is that this is also very likely in a busy environment. However this is not something you would notice at a heli event or most airplane events. If your system has a 1 mile (5,280 feet) working range and you are less than 1000 feet from your model you are unlikely to notice a range reduction. This would be of primary concern to glider pilots who fly high and far. So far, none of the glider events have gotten anywhere near big enough to likely cause band saturation.
Now, if you were to host a glider event while the IRCHA event was going on you could see a problem. But that is not likely to happen.
So the consensus is that latency creep is the most likely thing we would see. Unless someone is watching the RF band and doing time correlation of band state to model in the air time, there would be no way to peg the issue to band saturation. Would it be enough to cause a crash or a mishap? Could a latency jump from 10 ms to 100 ms be enough to case a problem on a 3D heli flight or a high speed pass of a jet? Again, hard to say with no one monitoring the situation.
So that is what I have learned so far, but would be very open to other thoughts or opinions.
In my opinion, large event organizers should consider setting up some kind of band monitor/recorder to see what is going on. The low cost versions may not have the absolute precision desired but they might be good enough to see if further study would be merited. There are probably less than 10 events in the USA that are large enough to merit this, but it might still be worthwhile.