We shouldn't throw away the physical paper copies just because something has been scanned. In any large collection of scanned documents, you are going to get scanning errors/etc – you can catch a lot of these with quality control, but some will always slip through the cracks. (And, volunteer efforts to scan old computer documentation often don't have the time and resources for the best quality control.) If the original paper copy is still in existence, someone can always go back and rescan it again later. If all the paper copies are gone, that is impossible.

Where are you going to store the paper? And keep it from degrading?

Paper is in fact pretty durable. A lot of History came through millennial papers.

I’m not saying that computers manuals will for sure stand the test of time for a the thousand years to come, but as long as they are protected from fire, they are probably a more reliable storage than a digital scan which will need humans to backup them regularly if they want to survive their physical media.

Sure digital backups, especially of PDF files, are easy and cheap, but they require an active human effort to keep the system alive.

Obviously I'm describing the ideal situation, and sometimes practical realities mean we have to compromise on the ideal.

Still, as much as is practical, we should try to avoid throwing physical documents out even if we know they have been scanned.

The maximum useful resolution you can really get out of offset-printed magazines is around 200dpi. Scanners have been able to do 400dpi for many years, and have been stuck at the 1200dpi plateau for about as long.

Sometimes you just don't need a better scan, because you're just doubling up on data you already have.

There are lots of other issues that can occur with scans beyond just the raw resolution: misaligned pages (part of the text accidentally cut off from the scan), accidentally skipped pages, colour illustrations scanned in black and white, etc.

With a sufficiently robust QA process, you may be able to catch most of these issues. However, you can never reduce the probability to zero, and people scanning old computer manuals don't always have the necessary resources to implement the most robust QA process.

There's also some types of papers and inks that don't scan well. Omni Magazine used metallic inks in some of their issues, as did many tech magazines of the period, probably because it looked futuristic. These don't scan well and even if the content of the page is preserved, the look isn't. At which point it becomes as question of how much of the value of storing artifacts of the past is in the presentation in addition to the content?

I'm not part of this scene and I'm not a collector, so I my dispassionate perspective is that at some point the cost of quality insurance is greater than the benefit. For a very important issue of a very important magazine, of course, keep it around. For a less important one, maybe retaining 95% of the data at 5% of the cost is good enough.

For many people there is a cost associated to storing items, be it old furniture or paperwork.

That's a question for folks who don't value original copies.

Meanwhile, mint-condition, shrink-wrapped Magic the Gathering cards sell for $$$. Same goes for other collectibles and real art.

Whether something is a collectable depends on whether someone (ideally several someones) is willing to take it or pay for it ... apparently in this case, they weren't.

Comparing a periodical to a mint shrink-wrapped MTG card isn't really a fair comparison. Even back in the day, most libraries didn't keep periodicals, they were "scanned" (photographed) onto microfiche at the end of every year or two to save space. Some stuff just takes up too much space to store, or isn't practical to when the information is of limited value or already present in another form.