> Every CEO I’ve worked under—no matter how compassionate they were on a personal level—has had a policy of absolutely crushing voices of dissent.
I think there's a really interesting confluence of factors here.
(A) CEOs tend to be psychopaths and/or narcissists -- people who care about others, or who don't have a bottomless need to prove their own self-importance simply don't become CEOs (or don't succeed at it). So of course they crush dissent: it is antithetical to their whole psyche. That comes from their weak, fragile egos, not any real "leadership" traits. That being said ...
(B) Luckily, having a singular, narrow-minded focus is sometimes good for a company. And the times when it's really needed, this narcissistic ego-fragility of the CEO happens to be beneficial ("wartime CEO"), at least for those companies that survive.
(C) Survivorship bias eliminates the cases where the singular, narrow-minded focus driven by the CEO's narcissism actually completely destroys the company. You haven't worked under them because they destroyed the company before you applied. Darwinism does the rest: good companies survive long enough for you to work for them because the narrow-minded idea the narcissistic CEO focused on happened to be the right one at that time. The company looks strong, the CEO looks smart, and then they exit to another CEO job and completely fuck it up, because they've never actually been smart, never been a good leader, never had all the answers -- they just got lucky, and the lucky ones are the ones who are around to be seen and worked for.
(D) When it's "peacetime", not "wartime", the narcissism of CEOs is slowly eating away at the company, corrupting it from the inside out, but it doesn't matter so much, because the company is generally doing well. They can hire and fire whoever they want; their "singular focus" is just ignored by the people who are actually doing the work, and the company plods along, being productive despite the cancer slowly growing inside of it.
We forget the bad CEOs, misattribute random success to good leadership, and ignore their parasitism when the company is doing well. We praise them when they save a struggling company, even if the company is only struggling due to their past ineptitude. And so we perpetuate the cycle of CEOs generally looking like strong, visionary leaders, when they are nothing more than narcissistic parasites who occasionally force a struggling company to randomly do exactly what it needed to survive.
Of course letting your employees pull in different directions is a better strategy for the long-term health of the company. But long-term health of the company is not what CEOs, nor shareholders, nor board members, nor anyone pulling any strings actually cares about. Of course this entire treatment mostly applies to publicly traded companies, or companies with an "exit strategy".
> Valve lets their employees pull in different directions
Valve is not a publicly traded company, and therefore its CEO is actually invested in its long-term health.