If I were engineering a light gun from a blank slate, I would probably first try Nintendo's method and see if I could meet requirements that way. It seems less

If I were engineering a light gun from a blank slate, I would probably first try Nintendo's method and see if I could meet requirements that way. It seems less timing-dependent and faster to market. It may also reduce the differences between NTSC and PAL versions of the product. Of course, blanking out the target for a frame makes the solution not scale well to a large number of targets. Presumably, one would perform something like a binary search (flash all targets, then flash half the targets if there was a hit somewhere, etc.) for a large number of targets.