Very interesting post, thank you.

Regarding the post by @silviu93 that you singled out, you are very correct in your assumption that something didn’t match.
That something mostly had to do with me traveling to Austria and surprisingly not being able to (1) use international roaming and (2) find some decent WiFi that allowed me to review/notify my team of my predicament until the post was 6 days old. When I finally managed to find some network, it was the spa’s WiFi, which meant I had very limited time to actually use my phone. After completing the review I rushed through the scoring, actually assuming the post would miss its payout. I have now checked back on the post and was surprised to see I was wrong on the last account (and I’m glad for Silviu on that account, since he’s been rather unlucky with payouts lately).
Nevertheless, your assumption was correct: most of the mistakes pertaining that translation batch were missed spaces or extra commas (which is probably the most common mistake in the Italian team: we don’t use the Oxford comma in Italian, but it’s easy to forget to take it off when the string is a simple list of words left in English). In general, though, I do agree with you that 6 mistakes should affect the other scores, too. 

I also agree on the fact that 5 out of 6 questions basically judge the same exact thing when a translation is well made. And frankly, I have never reviewed translations that were not well made, which I believe it’s the whole point of this project. I have never seen this as a learning environment but rather as a collection of already-skilled individuals. Sure, some of the translators have gotten better with time than they were at the beginning: practice will do that for everything and everyone. But the whole point in recruiting them was that they were already good at this.
I don’t particularly hate the new questionnaire and I actually like it more than its predecessor. But I find it redundant and in some aspects inadequate (why the gap in mistakes count? why give up on the major/minor mistake breakdown?). I had repeatedly voiced my opinion in writing  back when we were brainstorming, both in comments and posts, but was unable to participate in the vocal chat due to work engagements. It seems like the only thing that mattered, in the end, was the vocal chat though, and whatever opinion wasn’t voiced over there didn’t really matter. So I’ve just given up on the issue and simply use what tools I’ve been given.