Okay, the term everyone is looking for is
coercivity.
You can find a writeup of the concept here --
http://www.cybergenetic.ca/ebook/wrh08.htm#E70E236 -- though it's still not explicitly clear when it comes to the application to floppy media. (...and since this is what I'm going on, don't *completely* trust my interpretation, either.)
Basically, higher-density recording, by definition, requires the ability to write data at higher resolution; the bits (or more accurately, the signals that get converted into bits) have to get closer together. But as you put things closer together, they get pushed into a physically smaller area, which causes a couple problems.
In magnetic media, you have to worry about environmental degradation - *and* the stray field from the recording head. Making the media harder to coerce solves both problems - it stores a stronger field when it *is* coerced, and the higher-powered recording head needed is of the same or smaller physical size - if you don't have to crank up the power *too* far, you can benefit again under the inverse square law. (I'm not sure if that works out in the floppy case, but I thiiink I've heard the tale told that way.)
So, when you shove a HD disk into an HD drive, it sees it's HD, and writes at the appropriate power. Shove a real DD disk in an HD drive, and it probably drops the power to whatever level is required - but even then, you can *still* have some issues regarding head alignment, depending how the particular drive was built. (If the same HD-sized head is used -- which I gather is somewhat common -- it might, for instance, erase a swath *through* the area a DD head would contact, resulting in data that *it* can read fine... but pop it in a 'true' DD drive, and that might see bits that are 'half one' and 'half zero.' I've had something that fit this model happen in the past - when in doubt, zero/format your media on the machine with the *true* DD drive before trying to cart something back and forth.)
So... now let's look at what happens when you try to 'abuse' media from one density to the other. Punching the hole in a DD disk to make it HD means you'll be using lower-coercivity media with a higher-powered head... So the signal should saturate the media nicely, but it might 'bleed' to adjacent areas, and/or injure the disk. Oh yeah, and some sort of field is applied to *read* the media, too, so there's a chance that 'weak' field will look a lot stronger to the more-easily coerced DD stuff.
In converse, covering the hole on an HD disk to 'turn it into' DD means you'll be trying to assault the higher-coercivity media with lower DD-powered writes on either sort of drive... *in concert with* the risk of the same head-width/alignment problems above. Both the HD and DD drives are going to think it's real DD media, and respond accordingly. At least you shouldn't have to worry about reads flipping bits...
Before you freak out too much, realize that the floppy industry, 197x-2003, is much like the CD-R industry today. While it took a lot of R&D to found it, tons of manufacturers cut corners when it came to providing the cheapest media - and that same cheapest media was often what your commercial software came on, since a few pennies of difference meant a few more pennies of profit. When HD drives first arrived on the scene, plenty of manufacturers probably kept on using the same media while charging the new 'HD' prices. Meanwhile, as DD hardware dropped off the radar, some of the last DD disks probably were (and may still be) HD media in the DD casings. You never know, especially without independent laboratories testing this stuff. Meanwhile, plenty of "real," 100%-to-spec HD and DD disks have probably failed in the past decade for wholly unrelated reasons, be it other mistakes made in emulsion longevity, the ravages of potentially invisible humidity or mildew, stray magnetic fields, misaligned drive heads writing tracks readable on one drive but no others... All the *other* things that make floppies unreliable for archival. So don't go blaming *every* bad disk you had in the past ten years on some of these games manufacturers were thought to have played.
Now why am
I trying, when there've got to be smarter people afoot? ;-)
Edit: One or two boneheaded word-left-out/wrong-word-left-in mistakes fixed.
