Update main challenge reference input files with correct ionic strength

[andrrizzi]

In this PR:

• Updated the files for the main SAMPL challenge reference calculations (the files for the SAMPLing challenge are untouched). The protocol is exactly the same as the old one. The only difference is the number of ions in the solvation box.
• Fixed a typo and added a short description of the buffers composition in host_guest_description.md.

@jmichel80, @Steboss a heads up that the buffer used for the octa acids experiments was made by mixing Na3PO4 and Na2HPO4 so the experimental ionic strength should be about 41mM, according to my calculations. If I remember correctly, you mentioned you were using an ionic strength of 60mM in a previous PR.

[davidlmobley]

Thanks, @andrrizzi .

@mrshirts - these should be final files.