The problem is that NASA's patents may enable a subsequent technology that is also patentable that NASA will then need to licence. By holding patents they can get negotiating power on further technological improvements.
My least favorite part about the traditional ways are that they tend to be very poor at estimating FTE/WYE needs for a body of work, and will more often than not err towards over-estimation. This ends up necessitating the creating of busy-work to make use of all assigned personnel. Mostly by adding extra checks and documentation that many would agree are not very beneficial.
It's hard to be lean when you're given 15 more people than you actually need, basically.