This is slightly different from What does author order indicate?. Our group has debated when would it be appropriate for a student or a post-doc to be considered the senior ie. the corresponding author.
As an aside, for CV building purposes, would it be more beneficial to be the first author or the last author? Does it matter if one is an academic vs. in the industry?
Assuming that you're not in a field where it's "strictly alphabetical order" (like economics), the question of which position is more important depends on what stage of your career you're in.
If you are a beginning academic—a PhD student or a postdoctoral fellow—then the first-author publications are most important, as these will show you taking an active and leading role in your research. As you move up the chain, however, and reach more senior positions, having the last-author credit becomes more important, as now you're showing your leadership role in directing projects. You don't want to be stuck in the trap of being a "junior" partner in research collaborations, with the senior PI getting all of the credit for the work.
In my field (Computer Science), I usually assume that the first author and/or the "corresponding author" is the one who "did most of the work" and the last author is the one who secured the funding to do said work. In some cases the latter does not exist (e.g., if a publication is not funded by a research grant/contract) or these are both the same person. In my opinion and experience, the author with the greatest technical contribution should be the corresponding author. This is because most inquiries to the corresponding author will be technical in nature. If I were interested in contacting the authors of a paper for a non-technical reason (e.g., an inquiry to team for a competitive proposal) and I were unfamiliar with the authors (which is unlikely), then I would do a quick Internet search to determine if any of the authors are advisors of the others and contact the most senior one in terms of academic rank.
Therefore, I think it is perfectly appropriate and even desirable for whomever provided the greatest technical contribution to be the senior author. This may vary by discipline and/or country, however.
A few points.
Most(?) PhD students don't stay in academia. If they a planning on leaving academia for some other profession, does it really make sense for them to be a corresponding author? Once a PhD student has left Uni it's hard enough to get them motivated to write any papers, never mind answer future questions on it! Ditto for RAs, not all RAs stay in academia.
Let's suppose a PhD student does stay in academia and goes on to do an RA. It's unlikely to be on the same topic, so they won't be able to keep on top of the subject.
As @aeismail mentioned, I really don't pay attention to who the corresponding author is. If you are doing a PhD with a well known researcher, anyone who looks at the paper will automatically assume that the senior person had the original idea. The junior author gets to dispel this "myth" by giving really good presentations or by "author order".
