Are you saying that the Timeguard one in my photo cannot be stacked indefinitely (leaving aside the weight of adaptors causing them to pull apart *)? Or is there some topological factor that I have overlooked? The stand-off between the pins of the plug part and the earth pins of the sockets would seem to be enough to allow room for the various appliance cables to avoid the plugs on the next adaptor of the same design in the stack.
Goodness knows where I got that particular adaptor from or where I would buy one or more like it to try multi-stacking, even if I had the inclination to do so, given that I would never actually stack them, even if it was theoretically possible. For providing one or two extra sockets in a wall socket or a socket bar, an adaptor is fine, but if you need any more, either daisy-chain socket bars or (even better) connect two socket bars to separate sockets of a dual wall-socket.
I'm wondering whether the blanket statement "square pin adaptors are deliberately designed to prevent stacking" is a bit too restrictive, and should say "some square pin adaptors..." referring to those with no socket on the rear face on the opposite side to the plug pins. Is the example that I've got an exception to the rule?
Maybe there *is* a reason, in which case why not explain it rather than saying "get back to us once you've played with a pile of them" as if it should be obvious to me.
To get back to the original question in the thread title, was there any reason why the pins are different shapes between older unfused plugs and modern fused plugs, rather than the pins just being arranged in a different way? Was it because rectangular allows the springy part of a socket to make a wider contact area rather than a point contact, lessening the chance of localised heating on a tarnished pin?
(*) And the pulling-apart consideration would presumably apply equally to round- and square-pin adaptors, whereas we are talking about a distinction between the two: that round allows stacking and square does not.