I really like your counter argument, focusing on the process instead of the output. However I have to ask: isn't the end goal of submitting a manuscript to get recognition for it? Do you genuinely hope every time that your submission will be mercilessly torn apart and possibly rejected instead of getting the recognition for the work you've done?

Truly I have seen article reviews ruin a budding research scientist's career. Of course one could argue that other problems contribute to finding themselves in that position.

If there is something worth tearing apart, I would much rather it be torn apart than published. Recognition for my work is not what scientific articles are for, right? When I read a paper that has unaddressed problems, it's a waste of time - possibly years of my time if I base my own work on that weak foundation. I don't want to inflict that on anyone.

I would think that if peer review ruins a scientist's career, it would only be through a lack of support from the advisor. If the review is accurate the advisor should support the revisions, and if the review is totally mistaken the advisor should support resubmitting elsewhere.

I think reviews tend to be overly critical primarily because reviewers are forced to find excuses to reject 9/10 of the papers that landed on their desk, most of which they weren't even interested in reading.

I strongly believe that review could be almost exclusively through positive feedback, with crowd-sourced minimal annotations in the margins, i.e. "the idea in this paragraph came from <earlier paper link>", and "this idea is novel; wish I thought of it", and "there is a strong analogy between this paragraph and <some other tangentially related idea>"

It is not my experience that reviewers are overly critical. The notes I have seen fall into three categories: 1) pointing out the nagging issue that I knew but avoided, 2) misunderstandings, which are annoying but show how the paper is not clear enough, 3) genuine insights which I would not have seen. I have never wanted these to be more positive, but I can see the benefits such a cultural shift would have. Phrasing the notes as positive feedback would benefit the happiness of the authors, but make the notes lengthier and increase the effort for the (volunteer) reviewers. I wonder if the editor could reframe the reviewers' notes to improve the tone before sending them to the authors?

At least in chemistry where I work, a rejection rate of 9/10 at the review stage would be very unusual. It is possible that reviewers in the triage stage are harsher but I never see these reviews.