I have recently acquired some new local writing partners, and we have just started sending our work to each other. They said they liked what I'd sent, and made a few suggestions, but really not as many as I would like. And when they sent me their work ... um... I didn't really like anything about it. I don't think I'm a stellar writer, but I like to think I know a thing or two, so my opinion counts for something. But I didn't like these writers' plots, the 2D characters, the cliches, the dialogue, etc etc etc So what I did was give tons of suggestions - but not all of them. I think it's important to give some positive feedback amongst all the red, because I don't want people to get discouraged. I figure if they send me their revision, I can slowly give some more feedback, instead of doing it all at once. Then I was thinking - what if others think MY writing is crap. Wouldn't I want as much feedback as possible? I don't mean to sound all high and mighty here. Does anyone know what I mean or found themselves in a similar situation?
I have an approach to reviewing that's partially based on avoiding that problem. Whenever I review something I start by imagining it will be perfect (i.e.: beyond my skill to review it). Then, if I find what I consider a mistake, I lower my expectations to the level of that mistake. From that point on, I don't comment on more complex mistakes than that one. If I find a simpler mistake, I reset my bar to that point and, again, don't comment on subtler mistakes. That way my anchor on which is the first problem the author should correct comes from the text itself. Always from my imperfect point of view, of course. Additionally, when I find many instances of the same mistake, after a while I create a separated point saying something like "be careful with PoV switches, *brief explanation*"; from that point on, I mostly ignore further similar mistakes.
Also take into account these are drafts. My first draft was awful, my first draft of this novel is only a little better. Its editing that brings out the best in my dialogue and character etc, then the story becomes more interesting.
You have to take it on a case by case basis. You have to judge how sensitive the person is and whether you're prepared to take any fallout from hurt feelings. I think everyone can improve but if you deem the person to be beyond help then it might serve no purpose to dig the knife in. They perhaps like writing as a hobby and I think it's cruel to spoil their enjoyment. As for how to go about the critique, I'd only comment on the most glaring mistakes first. After that I would comment on other things only if specifically asked. Cogito and Banzai's comments in the link above are on the money.
There have been stuff I've reviewed where I just read through all the cliche rhymes, plots, themes, and endings. Sometimes the pieces as a whole are just so terrible I don't even bother going through each line saying what I don't like about it. If it's really THAT bad, I'd just move on. Sometimes I feel like messaging the person and saying, "You know what. Just stick to your day job." But like myself, they find an expressive form in writing. Which is something I don't have the authority to destroy. Just give them suggestions to think of something more unique. Some people just don't have that ability and in turn, come out with complete rubbish. I do say that I straight up don't like it, but there's a fine line with honesty and just being brash.