There is actually a substantial difference between the abundance of something in the earth's crust, (which can in general be estimated by statistical means by sampling the tendency of somehting to occur in typical common rock types), and its useful abundance in order to mine it (estimated from known economic deposits).

For example there are extraordinary amounts (in comparison to mankind's stock) of gold in the sea, but at an average concentration it is not economic to extract it.

As another example, copper is, on average, rather less abundant than you might expect given that it is the second most important non-ferrous metal in industry (after aluminium; on a par with manganese and zinc): but it is hardly more abundant than the "rare earth element" neodymium, and rather less abundant than such specialities as tungsten, vanadium and zirconium. But it has a useful tendency to occur in large concentrations which are convenient for mining. This is because vulcanism tends to mobilise copper and concentrate it by a kind of fractional distillation. The same sort of consideration leads to certain elements having a strong tendency to occur together.

Another example, gallium, mentioned by OP, is really fairly common stuff, only slightly less abundant than lithium whose production is 1000 times higher, and more abundant than lead whose production is 100,000 times higher. However useful concentrated deposits do not appear to occur, and it tends to be produced as a byproduct from other extractions, including aluminium, zinc and coal. No doubt if it was worth it, production could be substantially scaled up.

Even more alarmism was recently attached to indium, only 5 years left some people said after low prices led the Chinese to stop production. There is some uncertainty attached to the abundance of indium, perhaps because we produce so little of it (75 tonnes a year). But it is around about the same abundance as silver, whose production is 100 times higher. It was amazing how the stuff came back into production when its price went up by a factor of 10. The wikipedia entry on Indium http://en.wikipedia.org/wiki/Indium has a more convincing explanation than I can give of what rubbish these claims of running out of such stuff are.

It is always philosophically possible that places we haven't looked will prove to be completely different from those we have looked already. But it seems more likely that the laws of statistics will continue to operate. If we have had past success in extracting, say, iridium, from zinc tailings (as I believe they do, along with other platinum series metals), and certain other standard locations, it seems likely that they will continue to be able to do so.

I read that the abundance of iridium in the earth's crust is extremely uncertain (probably a result of its tendency to be found in meterorite-derived material, an dthus less dispersed than most other elements). Wikipedia presents numbers from a variety of sources given at between between 0.001 ppm and 0.000 000 3 ppm, though going to the source of the latter number I think it should read 0.000 003 ppm; but the origin of that is in turn not clear, and might just be a mistake. The annual production of it is 3 tonnes, which is a great deal if the lower estimate of the abundance is correct, and hardly any at all if the former is - the former is about the same as gold whose production is 500 times higher.