The best guess at this point is that the force of the impact split the valve, allowing a rapid flow of gas through the dip tube, which caused friction heating on the thin lower edge.
A couple things to consider...
It is true that a high flow rate of gas will cause heating on sharp edges, protrusions, etc. due to the extremely high flow velocities that can occur. In general scuba valves are terrible candidates for O2 service as they have sharp bends, sharp edges and open very quicky, creating the potential for heating. But they are all we have, so we deal with it, handle them very carefully and open them very slowly, and keep the flow of gas through them as low as possible.
Removing the dip tube is not going to do anything to prevent a future occurrence of the accident that occurred. If what you suspect happend but the dip tube had not been present, the gas would have escaped past the threaded portion of the valve where the dip tube would have been. Consequently, with a high flow rate, the sharp edges of the threads would have heated as well. In fact, given that they are sharper and thinner than the end of the dip tube, and there are more of them, it is pretty obvious that they would reach a critical ignition temp sooner than the end of the dip tube itself.
Even if you drilled them out and properly radiused the edge of the hole, the gas is still going to pass over a sharp orifice on the way through the valve and it will also rapidly heat in a high velocity flow.
Even though you have the best of intentions, you are probably increasing the risk and your liability by removing them from tanks you prepare for O2 service.