I’d imagine it would be pretty easy to design an experiment (although not necessarily to get good results!) where you build a few dozen detectors made up of a He3 or BF3 neutron counter and an antenna to pick up the EM pulse together with a GPS location/timer accurate to a few ns and data logger. Then space them in a grid a mile or so apart in a thunderstorm prone area. Leave them logging every strike for a season, then come back and process the results.
You can triangulate the position of the lightning strikes fairly simply, then although the neutron detections may only be barely above the noise level, over a few thousand strikes you may be able to build up reasonable statistics, using the delay from the EM pulse, giving not only the position (co-location with strike?), energy distribution, and maybe infer the height (ie do they come from the low level arc, or the jets/sprites higher in the atmosphere)