That's only true to a certain extent. Don't forget companies are mainly run by profits. That's what really ends up deciding how the company is run.
Do you think a "good guy CEO" would stop Microsoft from extracting billions of dollars worth of royalties for bogus patents from Android and Linux vendors? Of course not. If there was such a CEO, he'd be quickly shut down by the board. And if the board were "sweetheart angels" that would ask for that, too, they would be replaced by investors.
So companies are really more like a machine, and the people are more like cogs in it. The machine's main goal is to produce profits. If one of the cogs says "no, that's not what we primarily want!", it will very likely get replaced. It takes a tremendous amount of will power to fight against the machine's desires, and it's usually only strong founders that can do it (Steve Jobs, Jeff Bezos, Elon Musk).
But those are the exception, not the rule. And even then, it's more about delaying profit gratification, rather than foregoing some profits altogether. It's about them knowing that if they do it their way, the company would be even more successful and profitable in the long term, while the machine's forces demand the profits right away.
In fact, ethical behavior is just another knob in the machine that people running a company can turn either way in their quest to maximize profit.
Erring on the side of too much ethical behavior can certainly impact profit in the form of lost efficiency and market opportunities, but we have to recognize that companies' customers are people too, and people generally have a distaste for unethical companies, and a fondness for ethical ones, that may influence their decision do business with them.
So turning the knob too far to either side can have direct and indirect implications for profit, and it's up to the people running the company to calibrate that knob in either direction.
Oftentimes the courses of actions a company can take on any given issue will not be a simple sorted array of choices with an obvious straightforward relationship in the form of less ethics -> more profit, but rather a complex decision matrix of causes and effects for which the outcome of any single option will be impossible to fully predict, and investors can generally see that too. So the people running the company will usually have some degree of freedom to experiment with which direction to turn the knob on ethics, and how far.
> The only way ethical behavior affects profit is when it causes projecting the appearance of being an ethical company to become less expensive. To whatever extent that "people" prefer doing business with companies that they see as ethical, what they see is a company's image and branding, not their actual behavior. The balance is between ethical behavior and PR costs, not ethical behavior and unethical behavior; and that's even a gross simplification, because unethical behavior doesn't immunize you from good PR.
Maybe the effects of good PR can be less easily measured and felt than the obvious and immediate repercussions of bad PR, but the effects are certainly real and build up with time, so reasonable people can disagree as to what degree a company should pursue building up goodwill through ethical (likable) acts. Ethical decisions and behavior is the foundation upon which good PR is built. Even the most obviously opportunistic PR stunt must have some backing in an action the company has taken that can be construed by their audience as doing good.
Companies take on the behavior of the personalities of the people at the top. Nadella, like Gates and Ballmer before him, was selected based on shrewdness, and everything flows from there. It's simple. The company will continue to bear the culture born of Gates because of their board (at the top), and their HR department (everywhere else).
However, shrewdness is just one trait, and I think it's disingenuous to insist that we can reasonably ignore all other consequences of having a different human being running the company because their shrewdness makes them as interchangeable as planks on a wooden boat.
Personalities and management philosophies can't be reduced into some simple measure on a shrewdness scale, and the other differences between these real human beings that you don't seem to be taking into account can and do bring real change in how the company behaves and operates. Maybe not a lot of the change is concentrated on the axis that you've cherry-picked, but that doesn't mean we can dismiss all the other change that may have resulted from this change in management as "next to nothing".
They in fact back fill the role with someone they think should be in charge of the future of the position (which is very often in a radically different direction), or more often than not eliminate the role altogether.
By your estimation, when Steve Balmer retired, the board would have hired another Steve. No, that's not the case. They hired Nadella who is different.
Microsoft still sells (almost) no hardware, PC sales are dropping like a stone, corporate sales are lagging (my Fortune 150 switched to Windows 8 a YEAR after 10 was released), and Microsoft has to monetize seats of Windows somehow. What's left? Collecting and selling user data and advertising. This sort of thing has to have been approved by Nadella. Is this the move of a "kinder, gentler" CEO who just loves open source and freedom? No, it's a move by a shrewd person who is desperately trying to keep his company relevant in the face of phones, tablets, clouds, and social media.
I think whether a company can change is the right question, because it stands to reason that if a company can change, then given enough time, it will change. How much time is "enough" for any given company is anybody's guess, and as stated in my original post, I'm certainly not claiming that in this case Microsoft has in fact changed for the better. But to rule out that possibility completely because of what they've done in the past strikes me as irrational.
Companies are not people. Companies can't have nor lack a moral compass that tells them right from wrong.
Companies are run by people. These people are the ones making the business decisions and doing the work that enables the company as a whole to do ethical or not-so-unethical things.
The people running these companies can and do get replaced. And when enough of them get replaced, the past actions of a company will no longer be representative of its potential tendencies from that point onwards. This is very simply because a new set of people will be the ones making business decisions and doing the work, and these new people will become the real moral compass behind the company's decision making.
And let's not forget the fact that even people can change. So what's stopping them from changing while running a company, which would thereby be reflected in the actions of the company itself?
All of this change tends to happen slowly and gradually, but change itself, for better or worse, is inevitable as long as a company remains solvent.
Of course, I'm certainly not suggesting that I know for a fact that Microsoft itself has turned an entirely new leaf. In fact, I find all of their recent actions toward limiting user freedom and privacy as disgusting, exploitative and unethical as most others in this thread.
My point is, however, we should try to judge each individual action of a company on their own merit or lack thereof, rather than always viewing them through a lense tinted by some extrapolation of their actions from 20 years ago into some fundamental measurement of the company's intrisic goodness or evil. At the very least, we should try to make an effort to calibrate our lenses to weigh new information more favorably than old, because otherwise, even when the company itself may have long moved on, our opinions of them will forever be stuck in the past.