The article does mention that one of the big issues Google encountered was logistics.

It's one thing to scan a few dozens/hundreds of books. It's a completely different thing to do it for all books. Assuming you'd want to digitize 100 million books in three years, you'd need to process 91324 books per day, or roughly one book per second, assuming no breaks and 24x7x365 operation.

As the article said, Google poured hundreds of millions of dollars on this, so I'd wager it's not as trivial as it sounds.