I’m in the fortunate position that most of my work is with universities. Not only does this allow me to meet a lot of interesting people and review a wide range of interesting articles (one of the better parts of the job!) but it also means that I can always rely on payment. And because I can rely on payment, I can afford to have a much better relationship with the requestor.
There will always be some parts of an article that are hard to understand, exotic terminology might be used, and of course, it’s vital that the Methods and Results sections are interpreted correctly. In cases like these, I find it best to consult the author – I typically highlight any points of concern when preparing the draft version and return it to the author for their opinion. I can then take their suggestions and complete the work.
Of course, after completing the work, I’m always open to further comments and suggestions from the author. My aim, after all, is to help get work published!