I had no idea how close we came to losing Beowulf. I wonder how many other Old English poems have been lost.
In the generations after, only the manuscripts that impressed, or writings that were the most celebrated as literature, were preserved or duplicated. Those not held in high regard were sometimes recycled for paper, or ultimately lost when the last copy fell apart. Cultural biases play a role too. There wasn't much interest in preserving pagan legends in newly Christianized England, and so much of the Anglo-Saxon literature we do have is from the later period and usually utilitarian or Christian in nature.
The scale of the loss is clearer further south. We have book lists from the Roman and Greek classical era. The playwright Aeschylus is thought to have written about 100 plays, of which 6 survive. Julius Caesar and Aristotle were prolific authors, and celebrated individuals almost continuously since their deaths, and yet the majority of their works did not survive.
Gutenberg's revolution hinged not only on moveable type, but on rag-based paper (wood pulp wasn't widely used until the 19th century), and reduced the cost of a book by about an order of magnitude, possibly two. Far less expensive than before, but still phenomenally expensive contrasted with a paperback or digital download.
That said, yes, the production, reproduction, preservation, and distribution of books was very highly limited. Much official practice relied strongly on individual memory and recollection, with minimal recordkeeping via media. See the use of tally-sticks used for recording and reconciliation of debts (and the burning down of Parliament, eventually), as an example.
As with most works of antiquity, a large part of the present ascribed value is due to the failure of most such works to survive.
There's also the historiography of early records, and how and in what form(s) early documents did survive, where and when they did. Often it's only through quotations or citations that any record exists at all.
I've found it delightfully ironic that one of the more tedious forms of online debate, popular with top-quoting users of Usenet and email lists, of citing passages and responding to them directly in-line, was apparently pioneered thousands of years ago, and that it is often the harshest and most strident critics of earlier works who are to be credited with even a partial survival of them, by virtue of their anacronistic Fisking.
Peter Adamson's History of Philosophy podcast addresses these aspects in several episodes, largely on earlier philosophers.
Take Offa's Dyke for instance, very little is known about its purpose from historical records (though we can guess). A structure which ran the length of what is now the English and Welsh border.