> Why the court concludes that that "would not be substantially different" is unclear from just this footnote.
It's because of two primary points made elsewhere in the ruling.
1. Copyright law tolerates lending by libraries in the case of print books because those books eventually wear out. Digital copies, on the other hand, arguably do not wear out. Therefore, the court does not think that what is tolerated for print books should be tolerated for digital books. It does not address the fact that print books can be lent out hundreds or maybe even thousands of times before needing to be replaced, whereas some publishers are treating e-books as "wearing out" after about 25 reads, at which point the library has to renew its license.
2. Publishers have established a very profitable licensing arrangement with libraries for e-books, and CDL undercuts it. One could argue that if CDL had been an accepted fair-use exception from the beginning of digital lending, such a market would have never taken off, in which case the "CDL undercuts the market" argument would not have had the same weight. But here we are, like it or not, in a time when most of the major publishers have established these licensing terms, and so the court observes that a market exists that can be undercut.