I got a mechanical engineering degree and specialized in materials science. I also worked with both heavy iron/steel at an oil/gas company, as well as with composite materials (carbon, aramid (Kevlar [R]), and glass). Not taking sides here, as there is a bit of truth to both. My opinion is that a statement like "It is a well known FACT that aluminum doesn't hold up as long as carbon" is akin to a statement like "gold is heavier than feathers." It all depends.

Carbon fiber in the practical sense as applied to bicycles is a composite material comprising carbon fibers in an essentially plastic (resin) matrix. The carbon has high strength/weight and strength/stifness ratios, but you wouldn't want a solid beam or tube made of it (which would be nearly impossible to construct). The matrix itself has poor mechanical properties, with very low compression and tensile strength as compared with Ti, Steel, and Al. However, the plastic matrix is just strong enough to orient the fibers in such a way that when the frame tubes are placed in bending, half the carbon fibers are placed in tension, which is where carbon fibers excel. So long as that plastic matrix is intact, the carbon can do its job. However, if you impact the matrix (such as in a crash), or you fatigue the matrix, you weaken the plastic matrix, and no longer can the matrix do its job to position the fibers, and no longer can the fibers then do their job.

Aluminum alloys, on the other hand, don't have nearly the tensile strength and stifness of carbon (on a gram to gram comparison). But aluminum is an essentially isotropic material (same properties in all directions), and is ductile enough to withstand a pretty serious impact without failing. Pure aluminum fatigues to almost zero strength over time, but the way in which it is engineered, the useful life of the frame will generally far exceed how long you want to own it. Also, the various Al alloys can minimize fatigue. Steel, by comparison, will fatigue to about 50% of its properties, and it reaches a limit, so its a bit safer. Carbon fiber composites have the trade off that they are more prone to impacts and other unexpected stresses.

If you look at the charts of material properties, you will indeed see a variety of strength/weight ratios, stiffness, density, fatigue life, etc., but there is enough technology to engineer each of these materials to perform in a desirable manner. CF is the newest technology, but its commercial appeal has generated enough $$$ in relatively recent years that bike mfrs can now construct reliable and generally affordable equipment from it.

One type of test your friend might perform is an impact test, where you subject specimens to a weight dropped from a known height. You might use a sandwich beam - having Al, Steel, Ti, or carbon skins, and a 1" thick core. You might also subject such a specimen to slow (non-impact) bending. If you use the same thickness for each material, you will get drastically different results. However, in practice, engineers account for the difference in material properties to engineer parts with different cross-sectional areas, facing thickness, etc. That is why so many frame materials are successfully used in bicycles today, and none have yet to become obsolete. Thanks to engineering, the differences in the end products are more subtle, and people can choose the ride characteristics they prefer. All of these involve trade offs.

Ill shut up before I ramble too much.