There may be a disparity based upon how inflation is calculated. Plus or minus 40% after an interval of 75 years is not unreasonable for the purposes of an "about this big" comparison.
A Paasche index will understate price inflation, while a Laspeyres index will overstate it, due to demand elasticity. Governments may change the formula used to report changes in the money supply, using different measures like M1, M2, M3, and MZM, along with the average velocity of money. This obscures the amount of value intentional inflation extracts from the productive economy, at the cost of mathematical accuracy.
You could buy exactly one Manhattan project with 1939-1946 dollars for $2billion, of which 90% was production, and 10% R&D. You could buy exactly one Apollo Program with $25.4billion in 1961-1972 dollars. One LHC costs $9billion 2008 dollars. Tevatron was a bargain next to that at only $120million 1980 dollars.
Put into that sort of scale, with $19billion in 2014 cash, you could do any one of the following:
Build an entire city designed for 50000 inhabitants supplied entirely by renewable energy sources.
Build an entire archipelago of artificial islands.
Bury an Interstate highway under Boston.
Build 3 international-class airports.
Build two subway lines under an existing major city.
Replace 5 long-span highway bridges.
Buy 6 Human Genome Projects.
Replicate the Channel Tunnel.
Does that bring it into perspective?