Technically it was accomplishments that were cited. And in the case of the accomplishment of building Linux, that was done so prior to Linus receiving his CS degree.

That said, university was noteworthy in his case as that is where he gained access to Unix systems — something a regular person at the time wouldn't normally have access to otherwise — which ultimately developed his interest in creating Linux. Access to machinery regular people don't normally have access to is the value-add.

Which is the contention around modern CS degrees. What machinery, as it pertains to CS, can't you reasonably access today without going to university? Even state of the art GPU/TPUs, which might be the closet thing analog to Unix boxes of Linus' time, can be rented at home for a tiny fraction of the cost of attending a university.

> Purely self-taught yet outstanding engineers are the exception rather than the rule.

There is probably no such thing as a purely self-taught engineer, at least not alive today. Formal educational resources have been so easy to access over the past many decades that there is no advantage to being purely self-taught.

Before the advent of the printing press, you probably had little choice but to be purely self-taught. Access to formal resources was reserved for the lucky few. But those days are long, long, long behind us. I don't suppose the distinction between "self-taught" and "not self-taught" is even real anymore.