There's a tendency among some critics -- and I've done this myself -- to review books in part by paying attention to what the authors say about their works (whether through footnotes, interviews, and other sources not always directly included in the text under review).
As I've said, I've done this myself (e.g. in some articles in my first book, Black Lightning) and so know first-hand what can be the limitations of this approach, even as it can be a viable approach. Part of the limitation occurs when the reviewer seeks to deconstruct the path between the poem (if a poem is being reviewed) and then the texts "about" the poem, It's rarely possible for a third-party to reconstruct fully what happened between authorial intention and the result of the work. Where the critic fails is when s/he automatically concludes that the work, consequently, fails because the dots can't be connected in some adequate fashion.
This would seem particularly important in something like poetry, which often transcends autobiography and/or authorial intention.
The process becomes even more complicated when the author says that there is a multiplicity of inspirations and sources to any one work. Because what the de facto deconstructionist then attempts to do is not just trace any particular source to the poem but also the (inter)relationships among those multiple sources.
And maybe this is a strong argument for looking at the work on its own. Not from seeking to erase the author (which is what has come mostly to mind whenever I've been faced with the issue of how to consider the author when looking at a particular work)....but because the deconstruction of how sources and result rub against each other is not always a source for critical credibility.
The exercise, of course, can be its own art form (and one of my favorite exercises -- hence my predilection for "engagements" vs "reviews"). But that, of course, is a separate -- and equally controversial -- story. Controversial because the fine line separates criticism and appropriation.