Some of them surely realize they're just slinging barely-if-at-all-justified BS, but I also think lots and lots of people are just terrible at reasoning about that kind of thing and don't realize how meaningless the numbers they're generating are. They're trying, they just suck at it and no-one's bothered to tell them (or seems to care).
Possibly programmers are more sensitive to this than most, and are reluctant to put forward "bullshit" numbers that would, if they did, in fact be accepted as reality by the folks "above" them. Meanwhile someone down the hall's being promoted for numbers that are even more a work of fantasy than those, and may not even be intentionally deceiving anyone.
Note that the sales folks don't sweat over how much of their numbers can be attributed to the people making the thing they're selling. Those numbers are theirs, period. "Did I sell that or did Feature X put it over the top?" fretted no salesperson ever.
Take a finance guy, for instance. You give him a set of rules and objectives, and there is no need to define what "success" is like, based on the output. Success is his ability to make money.
A software engineer may implement features A, B, and C in the product. She may even increase the performance of the pipeline two fold. But there is no way to objectively quantify the impact of such changes within the current work organization frameworks. These are tools for middle management to quantify a team's output, nothing more. Thus a good software engineer would get a nice bonus at the end of the year, and a compulsory, but meaningless, promotion.
Now, I'm not saying that promoting individualism is the way to go, nothing further from the truth. I understand that there are intrinsic differences between finance and software. Yet I believe that there are very few companies out there with the right tools to evaluate the output of software engineers, and recognize it accordingly.