The supposed decline of copyleft

if its use declines (provided this claim is true) is it because we care less about software freedom, or because there's less need for such a tool?

That's a false dichotomy, both in the assumption, and the conclusions. As John's talk pointed out, “decline” could mean a lot of things, and we often don't know which one people when they claim there's a decline:

• Is it a decline in new projects choosing a copyleft licenses?
• Is it a decline in copylefted patches to existing copylefted works?
• Is it a decline in copyleft enforcement?
• Is it a decline in enthusiasm of copyleft?

As for the reasons for such, should any decline exist, the potential reasons aren't two, as you suggest, but numerous. In addition to your two, there's also:

• copyleft violations are so prevalent that developers feel that copyleft is not functioning to protect their code
• Software freedom hasn't succeed vertically, but laterally, and as such there is some Free Software in every endeavor of software, so the pain of proprietary software is lessened, but far from defeated. Nevertheless, developers has enough freedom they don't feel the urgency for copyleft protection that many once felt.
• The level of average education about what licenses do and how they work as decreased throughout the software industry, and as such developers aren't aware of what copyleft can do for them.

but the latter would mean we are now close to the moment we'll say: mission accomplished, unarm the weapons.

This is the issue I was writing about here, and what the second point above is getting at. There's lots of Free Software being written, but more software than ever in history being created, and therefore there is lots more proprietary software than there once was, too.