There is nothing wrong with fuses, small or big, mA or kA. All of them are fully predictable in whole design range, as specified in datasheets/standards,
if not abused.
Dave took a fuse designed to deal with 4In to 10kA and abused it by heating it to insane temperatures at 2In. Typically such fuses/fuseholders just crack and crumble from overheating, well these just popped gracefuly. If you take a "partial range breaking capacity" fuse (these start from "a" as per
IEC60269), you can destroy it exactly same way, by overheating. Mind this is stupid and most likely dangerous, these are fuses to deal with shorts, not overloads, and require overload protection. In residential environment much more often you can find "full range breaking capacity" fuses, (these start from "g" as per
IEC60269) and can work in circuits from 0 up to 10kA,..,100kA of prospective short circuit current, whatever design parameter is specified. One cannot overheat them when operated/installed as designed, the list of possible abuses is shorter but still quite long. There are various fuse standards around the world, "residential fuse" is quite a wide term, I am not sure about Australia.
Residential fuses and fuses in general are much more reliable than circuit breakers. I have never heard a fuse to fail. These do trip but always as designed, this is not a failure. As for tolerances - that is another subject and a matter of comparing datasheets. Not to be confused with discussion about a tolerance for abuse - this is yet another, but not very interesting subject. Especially in the context of handheld multimeter protection and by ab-using reputable brands fuse manufacturers as reference of such. IMHO.
If you want to take a look at circuit breakers "amazing unpredictability" - I can recommend bigclivedotcom YT channel. And no, these do not always fail open.
As for abusing circuit breakers - take a look at mikeselectricstuff near-death experience where he was "extreme overload"-ing RCBO.
Abusing fuses - EEVblog 1377 of course, but seriously, this is the job for Photonicinduction! How about 1kV across low rupture capacity 250V CAT V multimeter fuse, wouldn't that be entertaining?
OTOH, I'd love to see an abuse of a proper 10kA fuse in action, @mikeselectricstuff, do you still own the Destruct-o-tron?