Sure, if we assume that the total budget for the 17 national labs is $14B, that would imply that the average lab is a bit less than $1B/year to run. Hence $40B can run an "average" lab for around 50 years. Or am I missing your point?

funding is not even per year to year, but your point is still valid