I think a team of psychologists/psychiatrists should conduct a forensic study to look back at all the whackos that have committed mass murder shootings and see if there were any commonalities, tell-tale signs, indicators, common behaviors/traits, etc. Money well spent, even if it is inconclusive.
That sounds good in theory, but there aren't good records for that kind of thing. Crime data has been inconsistently collected and is still being inconsistently collected, although it's getting better.
For instance, 50 years ago somebody goes into rob a store. The store owner tries to jump the robber and the robber ends up shooting and killing them. The only thing that would have been recorded was the most severe of the crimes, the murder. There really wouldn't be anything in the crime data to point to the fact that it was part of a robbery as well, or the details of the events that lead to the shooting.
Yes, there would probably be court records from the trial and whatnot ... but can you imagine the effort involved in going to find all of that information and complete the data for all of those events?
20 years from now, what you are proposing might be available for about 30 years of data, but it would be hard to come by right now.