I also found a journal article [1] which says (p. 1748, my emphasis):
> There was considerable debate in the 1960s, during the gestation of the legislation that became the Copyright Act of 1976, about whether computer programs could, or should, be protected by copyright law. Although no one seriously questioned that source code forms of programs could be copyrighted as written texts, there were two principal concerns about applying copyright to machine-executable forms of programs...
So, according to that, the debate was primarily about whether object code was copyrightable, as opposed to source code. At the time, distributing software as source code was extremely common – indeed, configuration files were rare, configuration was commonly hardcoded in the source code, making compilation a necessary part of the installation process – which meant that source code being much more clearly copyrightable than object code would have been less of an obstacle to commercial software distribution than it would have been in later decades, when object code only distribution became much more common.
[0] Catalog of Copyright Entries. Third Series: 1972: Title Index. Books: July-Dec. page 3926 which lists "CILA Mark-1 system (casualty insurance logistics automated) source program listing. NETWORK DATA PROCESSING CORP" – https://books.google.com/books?id=4kAhAQAAIAAJ&pg=RA1-PA6 – note there are many other references to "computer programs" in that index, but it is sometimes unclear whether they are manuals or source code; this particular entry is rather clearly source code.
[1] Pamela Samuelson, "The Uneasy Case for Software Copyrights Revisited", George Washington Law Review, vol 79 no 6 (September 2011), pp. 1746-1782. https://www.gwlr.org/wp-content/uploads/2012/07/79-6-Samuels...