[bgalbrecht]

I didn't read the entire article from The Awl, I stopped after all of the errors describing how copyright used to work in the US:

The original US copyright laws, copied almost exactly from the British Statute of Anne (1709), covered books, maps and charts for a maximum of 28 years: 14 to begin with, plus another 14 if the author survived the end of the first term.

and talking about the Bono copyright act:

Had the various alterations to copyright law never been made, it would now be possible to quote, reprint, copy or reuse anything published up to 1981 or so for free.

and

The Copyright Act of 1976 created yet another terrible tangle by requiring authors of works published between 1923 and 1963 to file for copyright renewal. Absent the required renewals, those works have fallen into the public domain. And the renewal records from this mostly pre-computer period are a total mess, apparently. There is no authoritative central registry of copyright holders for these so-called "orphan works."

Prior to the switch to automatic copyright for life+50 years to match the Berne convention in 1976, the US copyright laws always required registration for a term, plus the ability to renew the copyright for a second term, it wasn't based on the author's life at all. The 1976 act did extend the term of existing copyrights from a maximum copyright term of 56 years (28+28 years) to 75 years. If copyright term had not been lengthened retroactively, everything written before 1955 would now be in the public domain, and only things between 1955 and 1983 not explicitly renewed would be in the public domain.

I wouldn't say that the renewal records from before 1978 are a total mess, although there is no authoritative central registry, the databases created from the digitization of the copyright renewal records by Distributed Proofreaders are good enough for most people needing to determine which printed works are in copyright. If Congress wanted a central registry, they could certainly fund the creation of one, I suspect it could be done for a few million dollars.

Anyway, since the author of the article totally didn't understand how copyright used to work, I question how accurate the discussion about Google's scanning and indexing books is.