I agree that the article sensationalizes the point somewhat. But at the same time, your dismissal of the concern because of the need retain knowledge, and hence the people having that knowledge, misses an important aspect as well.
The table shown in the article is tuned not only to weigh against losing people. It also -- and this is the really telling part -- is tuned to give preference to anything that increases the workforce.
Surely preserving the existing set of experts is not fostered by hiring on additional ones. That sounds to me like nothing more than empire-building. And I think that in the big picture, accounting for the possibility of future lean times, making NASA more dependent on personnel expenses must increase the possibility of having to let go some of the pile of experts that these preferences are amassing.
Not that anything can be decided based on this amount of knowledge, but isn't it also possible that they want to increase their workforce so that the remaining experts can pass on their knowledge while they are still there?
True, that's one possible explanation -- but it seems a stretch to me.
On the other hand, if the actual goal is not to increase the workforce size as such, why not just set up the criteria to reflect what your actual goal are?
That is, if your goal is to prevent "brain drain", why not specifically state the criteria in terms of loss of personnel possessing unique expertise. The criteria as stated would let us be perfectly willing to throw one Feynman under the train in order to save two cafeteria dishwashers.
The table shown in the article is tuned not only to weigh against losing people. It also -- and this is the really telling part -- is tuned to give preference to anything that increases the workforce.
Surely preserving the existing set of experts is not fostered by hiring on additional ones. That sounds to me like nothing more than empire-building. And I think that in the big picture, accounting for the possibility of future lean times, making NASA more dependent on personnel expenses must increase the possibility of having to let go some of the pile of experts that these preferences are amassing.