Resilience is really toughness rather than strength. CF is up to five times as strong as steel, but as noted has no ductility. CF will take a degree of damage, but it will either crack or delaminate. A crack may not propagate in a sheet, but it will be detectable. Delamination occurs within the structure, because of the nature of the layup.
Since CF is anisotropic we tend to align the fibres in the same direction as the forces it will experience. We use this to advantage, for example, in Formula 1 suspension arms that are incredibly strong and light for the job they do, yet weak enough you could easily break them if you stood on them. So for the less predictable world of bikes (etc.) we use multiple layers of carbon fibres in alternating or specific other directions to build in durability – insurance against those misdirected forces.
A bending stress in a CF tube will eventually cause the tube to exceed the strength of the epoxy; the epoxy cracks and delamination and splintering is instant. A point stress is similar, so if the epoxy is compromised the damage may be benign if the forces are low – recall that CF has essentially no fatigue limit under the forces it is designed to endure – or if the stress was created essentially at the surface only. However, a blunter impact can cause internal stress that leads to delamination between layers or between fibres and the epoxy, and this is why the material can appear unharmed at the surface but proceed to fail catastrophically later. Armouring a frame costs weight, so we protect CF parts from surface damage with rinky dink patches of tape, or gelcoat, or better, Kevlar.
The torpedo is almost wholly CF, but has Kevlar inside the wheelarches to protect against abrasion from the tyres, and thick gelcoat elsewhere. The WAW by contrast, and many of Mike Burrows' earlier creations, used carbon fibres with Kevlar as the binder. Kevlar provides for toughness that pure CF lacks but is slightly heavier than the usual resin.
Sorry, that's maybe more than you wanted to know.