I think your disagreement with Graham is over a definition of words. You seem to be saying that creation of value is the indivisible sum of efforts to (a) decide what to build, (b) build it, and (c) sell it. He assumes a and b can be isolated out.
In any case though, take any highly successful company. Imagine if all the marketing people and product/engineering/design people were suddenly segregated, and forced to work without the other. Who's more worried?
I'm not at all saying marketing people are useless, which seems to be the idea you take issue with. They provide value. But (I'm sorry) less than the designers and engineers, and measurably so, from successive thought experiments like the above. That's why they're paid less.
"The fact that value is subjective and immeasurable doesn't insult anyone who tries to make good things. I try to make good things too, the value I get from them is the sense of accomplishment that I've done a good job."
But by your argument, I can steal your sense of accomplishment by saying, who knows how much of your success was due to you? How much was due to the marketing team? 10%? 50%? 80%?
"But whether my "good job" translates into a $1bn unicorn is up to luck, not simply my effort."
On this topic I mostly defer to Hamming's point in You and Your Research. I'd say there's luck involved, not that it's up to luck. While it's true not everyone who can become Bill Gates does, some people outright can't become Bill Gates, while others can. The point is, this doesn't show that value is subjective.