He had him on his show once, it hardly makes his a master (much less "up there") which goes into the next answer...
A few reasons;
I find him arrogant when he declares other people's work "flawed" yet comes running into PC hardware review segments as a novice but declaring his reviews as the gold standard.
I find he makes mountains out of mole hill issues (mostly to extend his segments, more on that below) yet fails to really look into the issue to see if it even impacts users.
I find his videos far too long and can easily be condensed to five minutes if you remove the hair flipping, the constant use of "ums", and the repetitive dialogue. Now from a business stand point I totally understand the decision of long videos. Youtubers make far more money once their videos pass ten minutes but from the consumer side of watching them, they are painful to me.
The whole TPU issue with him a few years ago could have been handled with a simple statement from him (on air or on his site) instead he makes a 20 minute video so he can make money off of it.
Many of the GN web site reviews are video only now, I get the business decision of this but still sucks as I personally prefer written reviews.
That said the case reviews are excellent in my opinion.
I can understand part of what you're saying, but I also have some issues with your reasoning. The first reason you present is an outright logical fallacy. Inexperience does in no way render one incapable of telling if something is flawed or not. More difficult? Sure. Less likely to know what to look for? Sure. But that's statistics. Some people still manage to analyze things well despite not having much experience with those things in particular. An outside perspective can help you see things in a different light or see past bad conventions. For this to be a problem, you need to demonstrate that their takes are significantly wrong in some way. Also, hasn't GN been running PC hardware reviews for like a decade? Sure, that wasn't their focus in the early years, but calling them inexperienced at this point is rather strange. And whether it's Steve or other team members doing various parts of the work is immaterial to this.
Also, saying he fails to see if issues affect users - how would they do that? Run a survey? Wouldn't they first need to present the problem to do that? And if a bad design gets by due to not breaking anything, is that a reason to not call it out? I'd rather say they have a set of principles and adhere to them. That makes their reviews trustworthy and less biased than most, even if it also at times causes them to overemphasise small elements.
I agree it's a shame that their reviews are often video only, but I never found their written reviews particularly good, so it seems they're just better at audiovisual presentations than written ones. And that's fine, even if it would be great if they did both.
I don't think I remember any GN+TPU thing happening. So it can't have bothered me much.
As for video length etc...Meh. It is what it is. GPU reviews and the like need to present all the data, so they're stuck with that even if it's all very skippable. I skip past 90% of individual game benchmarks in written reviews too, so that's a moot point to me. The videos where they discuss and analyze things tend to spend their time on arguments and looking into issues, which I find worthwhile, even if I might watch the video split into three. A shorter version of those videos would just come off as hyperbolic and poorly argued, which would be far, far worse IMO.
Are they perfect? Obviously not. But among tech reviews today, they're still among the most technically detailed, principled and - crucially! - curious ones out there.