There is no 'i' in research

Why keep writing papers?  Who is reading?

There are reasons to write good papers I know - advancing research by exposing facts and thoughts, and even if no one is reading now, someday it may light another's way.
There are reasons to write bad papers too - practice in one's academic field, and adding to the 'productivity' count for career advancement.   
I can think of other good reasons to write papers - training others, engaging in collaboration, combating boredom, ...

If we were to focus on advancing society however, there would be much fewer papers.  If you got nothin' to say, don't say nothin'.  I mean say it in your notepad, say it to your facebook friends if you want.  In this scenario, there would be bad papers, there would be important papers, there would be meta-papers (that review others, that explain connections with past and future work), there would be popular papers (that court power/policymakers, money/capital, that seek popular understanding and support) , there would be the orphan papers still (no problem)

There would be a great division of labour, with some specialist writers, even multimedia specialists, you know, to make cool videos and sculptures and installations.  There would be more communities (there are some strong communities e.g. in some critical physics research where it's clearly more about the mission than the men.)  There would be growing wikis to organize the information produced.  

The current system sort of works but more community would make it stronger and faster.  It would also make a more literate public that can then contribute to research - in fact and in kind.

I think 'everybody' should know math well - in the same way masses know good literature, masses participate in religion, masses debate politics.  Instead, the masses are shut out of math and science by the unsexy image and horrible PR.
Instead, there is an ivory tower breached only very slightly by cool publications - Caltech E&S (bet you never heard of it), MIT Technology Review, Science, IEEE Spectrum, Valleywag (just kidding - that's Silicon Valley news, but actually that's the idea: imagine if 'research' had a valleywag?  I mean, what's everybody so damn uptight about?)  Popular books and blogs are also the right way to go.  All this generates interest which generates wealth. 

My guess is that once you figure out how to fund it - you know it's easy to fund current 'tech' news because there are gadgets to be sold - you're basically able to massively produce science media.  It can be done; people figured out how to sell the tradition of Jesus, so it's not that science is too boring or anything.  

Academia ought to be recruiting thinkers at a higher rate (everybody ought to be free to be a thinker.)  Then also it ought to be more connected to the lowly world outside.  I once met a great professor that claimed expertise in Mesopotamian archaeology and such.  Not a word of Arabic he spoke.  Not a trip to Iraq.  Dude, how can I take you or your interminable papers seriously? 

Let's talk about what scientific information you use and how you get it. 
Also, there's a great opportunity to make science/research/rationalism/math sexy in these African parts.  Let's talk about this too.  I mean, some steps are simple: put the existing journals online.  Have themes/topics for some editions.  Let someone write an introduction/editorial.  Open up for comments. 

Disclaimer: I have not written many research papers, and indeed may never do so.  I sometimes try to write about math (xinvogue.blogspot.com)  I admire people who share their enjoyment of math.