On her substack post Victoria Moul wrote "Most reviews don’t seem to be good at making me want to buy poetry books: ... perhaps ... because they all tend to be so positive. If everything is apparently wonderful, it’s hard to trust any particular recommendation"
I think reviews are more positive nowadays. In tone they're more like comments in a face-to-face workshop with something good always said, adverse comments being sandwiched in, muted. Maybe nowadays reviewers are much more likely to know the poets, or at least they've communicated online. Or maybe reviewers feel that poetry's in such a bad way that it needs all the help it can get. This positivity (or at least lack of negativity) is especially prevalent when dealing with bereavement poems. Can there be such a thing as a bad poem about Refaat Alareer? The Poetry is in the pity I suppose.
I try to self-moderate my write-ups. After all, half of the books I read are worse than the average book I read. That's the way I rate on goodreads. The reasoning can be extended - half of the poems in a book are worse than the average poem, and half of a poem's lines are worse than the average line. Atomising a poem in that way is tricky though - lines interact with each other (jewelstones need mountings), and a poem full of beautiful phrases may be a mere "anthology of lines". But many poetry books are made of poems that can be individually assessed. If reviewers believe in the concept of ranking poems enough to list (and quote from) a collection's best pieces, why not list and quote from the worst too? It gives readers a better feel for the reviewer's prejudices, and the poet's range.