Which brings me back to my question: Is it a bad idea to take big risks for big rewards in war? Or only when it doesn't pay off?
Generally I'm not going to argue with results, but we've studied strategy and tactics for two centuries. Wargaming it was invented in 1780 by Johann Christian Ludwig Hellwig, for Pete's sake.
And the importance of logistics was well understood even by Napoleon.
We can have a pretty good idea when it shouldn't have worked, even in many cases it still did by sheer blind luck.
And to return to your example of annexing Austria: Austria actually had a chunk of population who WANTED to join Germany. The Austro-Hungarian Empire had built most industry in Bohemia. When it broke apart, Austria was left as a backwards nation. It already had people literally starving to death in the 20's, for Pete's sake. It was fertile grounds to be infiltrated by the Austrian national socialist party, and it was. It also was guaranteed only by Italy, who had just switched sides on them, and their army was symbolic at best. And also thoroughly infiltrated by the national-socialists. It was a safe bet that they would fall to Germany in no time.
France had a similar but different political crisis, where the whole country distrusted the army and crippled it intentionally. It also had thoroughly obsolete doctrines. And Belgium had just broken their alliance with them, leaving their northern border wide open. Also, most of their industry was concentrated near the German border, and it was predictable that it would quit fighting once that falls. Yeah, that did them in more than the fall of Paris. And even then Hitler expected a longer war.
By contrast, the invasion of the USSR was a crap-show in more than one way. Ignored the logistic corps that actually wargamed it, was based on racial beliefs more than any actual intel, and the invasion plan drawn up by Halder ignored any strategic objective. It was predictable that it would fail.