You're erecting a pretty detailed strawman here. It's not that "raw dogging" code makes you a better developer; it's that specifically jump-to-definition and autocomplete make it too easy to create code bases that may as well be huge bags of functions and classes, and that skipping the mental work of organizing the code limits the fidelity of the mental model you can build.
Programming languages are complicated. Their standard libraries are extensive. Real world application are usually not trivial, because they often model real world processes, and the real world is messy.
Anything we can do to make writing software easier and more reliable, and reduce cognitive load, is going to benefit the software developers who are involved, and will make the systems better.
I'm sorry, if you can't jump-to-definition then you are wasting the company's time. It's something that all developers need to do and there's no reason we should be wasting time navigating a codebase when what you are looking for can be found instantly.
Same with autocomplete. I can type very fast, but autocomplete can type faster. Plus with the size of most APIs (even the ones built in to most programming languages) I have better things to spend brain space on. Is it list.empty(), clear(), empty(), truncate() or something else? With autocomplete I can find the function in want in 3 seconds (and read the docs inline, so I can tell that empty() doesn't empty the list, it tells me whether or not the list is empty) without lifting my fingers off my keyboard. Should I remember which is which? Maybe, but I don't care, and a jump between languages frequently enough that it's not worth the effort to keep track of silly things like that
That's a contrived example, but hopefully you get the point.
We're SWEs; the temptation to min/max ourselves is quite high. And it's not a bad impulse! I'm a Vim user; I've obviously invested a lot of time in editing efficiency (e.g. I'm great at looking up docs in the terminal). Oftentimes when I need to code I can't dilly dally. I also prefer a style of dumping out an implementation as quickly as possible to prove/disprove the idea, which of course requires a lot of editing efficiency.
But the way I've gotten here is by assiduously removing things that slowed me down. Sure popping docs in your editor is faster than switching to a browser, but just remembering is even faster. If your goal is really to not "waste the company's time" then you're putting the standard library, your dependencies, your app, etc. into Anki and memorizing. Since almost nobody (some people are and bless them) is doing this, I think we should admit we're fully in the realm of personal aesthetic preferences here.
And I take a broader view of the whole thing besides. I start from the perspective that engineers are whole people with histories, futures, goals, features, interests, and opinions. For example, Go wasn't built with autocomplete and go to definition. You're more or less arguing that Rob Pike should've been forced to setup VSCode. I think that's an express ticket to pissing off and burning out your engineers; just like I'd never micromanage Pike to that degree, I'd never micromanage you to the point where I'd force you to learn Acme.
The value we bring to our companies isn't just the speed with which we crank out code. The value our companies give to us isn't just a salary and benefits. We have more nuanced and complex human needs, and sacrificing those for extreme coding efficiency may provide short-term gains, but the long-term effect is pretty grim (ponder for a moment working at a company that actually cares about efficiency to the degree they'll micromanage your editing workflow).
