Yes, if it was just for a specific region or area the data can become perishable if not maintained. Chris Scott's sahara overland has a great country by country summary for most of what we NEED to know, along with info about the risks. The HUBB and Chris Scott's stuff are frequently cited because they are usually the freshest information. To widen the net statistically, NGO workers tend to go through the same shit holes we do, and probably in greater numbers.
In aggregate, for all riders, it would be useful because it would inform the probability of a particular event, just like they do in the insurance industry with actuarial analysis. The survey could just ask if anything bad happened on your trip, and if it did, in hindsight, do you think there was any risky choices you would now make different. For example, there is a particular road in Guatemala where lots of overlanders have been robbed, even one was shot. I narrowly escaped a similar fate through sheer luck, but if I had done my research, these experiences are clearly documented on advrider.
Not to take the fun out of it, because there is a certain amount of hubris and "damn the torpedos, full steam ahead" that is required to overland. A good deal of the things I have been told are dangerous or risky turn out to be complete bullshit in the end in my experience. I would still like to know, and to know if the information is just rumor or based in fact, and would be happy to contribute my experience. This is not just for risks, but more prosaic question like visas stuff, petrol, insurance, food, water, accomodations, etc might be good to compile rather than have to do google search and parse through all the answers. Then there are things like which bike (and if any probs, which probs), what gear, budget, carnet, etc that might be valuable information to have compiled rather than random. I don't know about anybody else, but I spend an inordinate amount of time trying to come up with the right answers for this stuff.