soooo Samuelson was right, before he was wrong?

what the hell?
Not at all. Economic predictions are made on the basis of the facts and policies in place at the time. Give me a date for a given quote and we can look up the economic situation then. Both the Marshall Plan and the GI bills are circa 1947 in terms of economic effect and the unemployment rate rose from 1.5% or so in 1946 to 4.0--4.5% in 1948 (the current data series from BLS starts in 1948, so figures for 1880--1947 are from Christine Romer's published work).
My point is that there is a huge difference between changing your predictions as conditions change and being inconsistent or wrong. Those who predicted hyperinflation in 2009 are clearly wrong, as the prediction did not happen and there is no good argument that any of the basic facts changed which would account for the discrepancy. Those that predicted that austerity would tank the British economy were correct and those who said the confidence fairy would unleash unprecedented growth were wrong.
Those who predicted the breakup of the Eurozone in 2010 had a good model at the time, the reason it didn't happen was the fundamental change of policy where the European Central Bank reversed policy and started buying sovereign debt starting in 2012.
I won a bet because I predicted in September 2008 that the Dow would bottom out at 6200. It was pure luck; I have no model that predicts stock price so I was using a SWAG. When I predicted at the same time that the unemployment rate would top out at 10.5% I was using a basic econometric model that just turned out to be a better model than the US government's at the time (remember the promise that stimulus would keep the unemployment rate from going over 8%?). I was off a good bit on interest rates and inflation rates, so I had to do some polishing. Deflation was bad, but not the 3--4% I feared.