Here is a real newbie question. If the current 1Mb blocks are underutilised, is there any evidence that doubling them to 2Mb will make any difference at all?
It is not necessarily the case that the current 1Mb blocks are being underutilized. To answer that question you would have to define what transaction have sufficient utility? How do we even define and identify spam? The answers to these questions are not necessarily clear, which makes the follow up question even more complicated. My answer is that whatever transactions the miners choose to include in the blocks have sufficient utility to be included.
This should follow real supply and demand for block space, not an arbitrary blocksize limit. The blocksize limit should be continued to be used as an anti spam measure which should exist significantly above the current average transaction volume.