Jason G, a couple of quick things spring to mind...
1) You can cut half of the names and paste them into a second column in your spreadsheet file (column header of "Name 1" & "Name 2" for example). Now you can use two variable fields for the "same data" in your merge, manually creating a 2up.
2) Create the file 1up, then use imposing software to make a 2up (N-Up).
If your copier RIP has has a VDP option (such as FreeForm on Fiery), then you only have to do the merge in InDesign using the names. The "boilerplate" or "template" does not have to be merged and this can be sent to the RIP separate and only RIPed once...then the RIP combines the 5000 variable merges from InDesign with the single RIPed template data (rather than RIPing the whole document 5000 times).
If you are doing it all using InDesign, then only do the data merge with the text names, as you wish to keep this fast/simple. Then after you have your merged document, apply the "boilerplate/template" data to a master page to automatically apply it to all the merged pages.
P.S. You may need to split the datamerge into separate groups, perhaps only feeding 1000-2000 names at a time into the merge (Datamerge in InDesign is not meant for "industrial" merges!).
Hope this helps,
Stephen Marsh