I think the author is pointing to a problem but not correctly identifying what the problem is. The problem is deeper.
If the problem were that we have socialized risk of innovation but privatized the rewards, that wouldn't really be a problem. It's a good thing to incentivize innovation.
The real problem is that we don't incentivize innovation.
Instead, we incentivize investment. People who already have money to invest take on small calculated percentages of risk to make back large amounts of money. Frequently the money investors are investing isn't even their own, and they are paid a significant salary to do so, meaning they actually take on none of the risk.
The negative results of this are twofold:
1. Often investors don't invest in innovation. It's equally profitable to invest in number-twiddling schemes that provide no benefit to anyone except the investor, and often that's easier.
2. We speak of companies as if they are single entities. But the reality is that executives and shareholders and of a company are the beneficiaries of any innovation within a company, while innovation typically comes from workers with technical expertise who are typically paid a relatively limited salary perhaps with some minimal stock options.
The way around this problem is typically to get into a company early enough that the line between technical worker and executive is blurred. But that avoids only the problems with lack of reward--it actually exacerbates the risk problem because early commitment to a company is a high-risk endeavor.
In the end, I think that a lot of innovators don't innovate for money anyway--I personally would rather see something I make change the world than make a lot of money off an idea that doesn't make any difference (or makes things worse!). As long as I have enough to live I'm okay. But it would be nice if we stopped pretending that our distribution of wealth is at all meritocratic.