Again, I don't think that's an apt comparison. The NAFCOM numbers used in that calculation don't include things like program support costs, contingency costs, contractor fees, etc. Those are usually baked into the NASA figures sighted though. Add on top of that, NASA will sometimes independently fund R&D related directly to issues experiences. SpaceX can be the direct beneficiaries of that research without having any of the costs on their books.

I'm confident SpaceX would still be substantially lower than NASA if they had, but we need to be consistent in the methodology or it just becomes the equivalent of cherry picking data to underscore an already valid point.