But when the government takes equity stakes, the right complains even more loudly about interference, especially if they are voting shares.
The purpose of the DoE isn't to invest to get monetary award, it's to invest to drive basic engineering forward. The government isn't as business and isn't supposed to derive a profit, that's the basic fallacy of the idea that the government's budget should reflect how CEOs run corporations.
Throwing money at research, science, and technology is pretty much how you get breakthroughs. As a said, there's no formula, breakthroughs are often serendipitous, unpredictable, and happen over long time scales.
I view technological advancements as exploring an infinitely large landscape of possibility. There are mountains of opportunity, but you don't know if a given mountain is worthwhile until you scale it, and you may fine nothing, or a local optima, in which you see a better mountain farther away. From the ground level, most of the mountains look the same, and the landscape is always shifting.
What is the best strategy for finding the successful mountains? Hire as many teams of climbers as possible and distribute them around the landscape far and wide. That's what capitalism is, that's what the startup economy is, it's lots and lots of people doing local hill climbing.
There are some mountains that are too big to climb, the gear is too expensive, and the time horizon to summit, too long. That's the government's role in basic R&D, to fund research and development at an early stage that is not economic, has little chance for ROI in the short or near term, stuff that even VC's won't touch.
Could any private company have built the GPS system all our phones used, if the government hadn't funded rocketry, satellites (most for military reasons?) You've talking from the 1950s until the 2000s when practical consumer GPS became a viable business. What company could have stomached 50 years of spending for ROI?
Much of the technology we have today is the result of government spending in the military, national labs, the space program, and state funded academic research. Often, the private companies are successful in taking the basic advancements and figuring out how to manufacture them, but it is usually the government that proved the basic design first.