scoring will not damage the bar.
However, two conditions can cause the bar to break.
the first is if the lower A frame bushing collapses, causing the housing in the frame to crack and allow the bar to move around and break.
another cause (rarely) but defective bar when manufactured.
My point was, to the regular joe, was it a nick, was it over twisted, was it bent, all anyone knows is it is broken.
So saying that a scratch or a gouge will cause it to fail like that is just a guess.
Yes a metallurgist could determine the cause but most of us are not.
Neither of these are true.
It's not a guess to say that a gouge
can cause a failure. That is a metallurgic fact. Now, whether a given bar in a given car failed because of a gauge is just a guess, without analysis a broken bar is just a broken bar. Now, if the crack starts out of a set of vice grip gouges, you can be fairly certain what happened!
Torsion bars carry about 90% of their load very close to the surface of the bar. Again, that's physics. That's why there are hollow sway bars, and why they only have to be slightly larger than their solid counterparts to have the same rate. Now, a small scratch isn't a big deal and isn't likely to cause a failure. A gouge, especially one that has a sharp edge and was made with a compression type tool, can absolutely cause a failure. The sharp edge creates a place where forces are localized and concentrated, and a crack can form there and propagate. The other thing is with compression, you can form a harder section of metal. Again, that part doesn't twist like the rest of the bar, which can cause hardening over time and cracks.
Torsion bars aren't fragile by any means, and some nicks and gouges can be dressed so there are no sharp edges and that will be fine. But a deep enough gouge can absolutely cause a torsion bar to fail, and if the damage is deep enough even dressing the damage out could still result in enough weakening of the bar in a localized are to cause problems.
I definitely wouldn't worry about every scuff, scratch or nick. Most will be no big deal at all. But that's a lot different from saying that the bars won't fail if damaged. They can, and they will, if the damage is significant enough. And that doesn't take some giant chunk of metal peeled out of the bar. Just look at the torsion bar chart, and see how a very small change in diameter makes a fairly significant difference in the spring rate of the bar- going from .87 to .89", a .02" change in diameter, is a 10 lb/in difference. That's an 8% difference in rate on that bar.