I don't know enough about this, but can you speak to the line
"particularly when the expected benefit is low."
If the expected benefit for (most?) patients of adjuvant chemotherapy is indeed very low, then it seems to me that the effect on a person's later ability to work should be taken into account by a patient.
I assume it's almost impossible for anyone to rationally consider odds in a situation like this, so most people probably think anything that might improve the chance of survival is a good thing. But, like the changes in screening recommendations, it does seem as if chasing after tiny improvements in prognosis is not always the best thing to do.
(I may not be understanding Xeni's take on this, natch.)