They should absolutely teach like this. At least for my personal amusement. Give them a test with one question like one of those listed above, and then grade based on how many answers the students can come up with.

That option has been removed as of version 24 (and possibly version 23?).

The amount of memory Firefox uses has always been a catch-all statement for bad behavior and poor performance, not a strict indication of the problem of running out of RAM (usually).

Personally, for a very long time, if Firefox crossed 1 GB of RAM usage, I could pretty much guarantee issues with slow tab switching, glitchy scrolling, the browser as a whole briefly hanging every few minutes, sometimes even bogging down my entire OS if it decided to saturate the CPU.

The degree to which this effect manifested could pretty much be exactly tied to its current RAM usage. I started to notice it at 700 MB, it was annoying at 900 MB, and at 1.1 GB it was completely unusable. This had absolutely no correlation with the total amount of RAM in the system. If I have 2 GB or 32 GB, it doesn't matter. Once Firefox crossed 1 GB, it was all over.

So when people say, "It uses less RAM", I hear, "It's less likely to reach a state that I have noticed invariably leads to poor performance all around", not "I just saved $300 on my car insurance"... er, "I just saved 300 MB of my overall RAM pool that I can use for other stuff".