Footnotes by a landslide

Nearly 3000 votes were cast, and 95% of respondents preferred footnotes, with the remaining 5% either indifferent or preferring endnotes.  The results are hardly surprising to any reader, but they are surprising when one realizes how the major academic publishers have all shifted to endnotes.  I gather the main explanation for this shift--since it has nothing to do with reader preferences--are primarily logistical and financial:  type-setting costs, the ease of converting to an electronic format, and the like.  But since I suspect the preferences of philosophy readers are little different from the preferences of other academic readers, there is a clear opportunity here for a publisher to lead the way.  I for one would give preference to a publisher that maintained a commitment to footnotes.  What do others think?  What explains the demise of footnotes?  Are there other considerations at work?