1. Yes
   1. Authoritative
2. No
   1. big errors
      1. content errors
         1. extreme values like inexplicably very large or zero
         2. glaring logic errors
         3. mathematical errors (i.e. equations)
         4. broken links (404)
         5. inconsistent numbers in proxy data. (we take data from somewhere and it's changed)
         6. grossly Invalid assumptions (proxy materials)
3. Examples
   1. aluminum scores better/higher than steel on energy efficiency
   2. A material that scores illogically well or poorly across the board
   3. one scale factor got applied twice in some cases -Ward
   4. URL in MSI spreadsheet links to inappropriate or broken page
   5. A value for <proxy material> does not match source value for same proxy material. *this needs to be further distilled
   6. Aluminum is a proxy material for goose down
4. Automated acceptance that changes made do not break the data.
   1. Fluid data is checked for tolerances
   2. Fixed data remains fixed