Let's start with her theory in its simplest form:
"A concept is a mental integration of two or more units which are isolated by a process of abstraction and united by a specific definition",with a "unit" being
"an existent regarded as a separate member of a group of two or more similar members" (ITOE, p11)eg: Two stones are two units
So in effect two stones are observed, from which the concept "stone" is formed, and then man's distinctive "conceptual" mode of cognition is up and running regarding stones. Well, that's all very well, but thus far described the process is all mental. Unless one can mind-read there seem few observable consequences we can take from this assertion. Fortunately, after some ambiguous hinting, a dozen or so pages later Rand provides us with something a little closer to what we've been looking for:
"The process of forming a concept is not complete until its constituent units have been integrated into a single mental unit by means of a specific word." (ITOE, p24)
Apparently properly formed concepts are vital to the "cognitive efficacy of man's mind" (ITOE p3), so we can assume that if a word is used without a properly formed concept behind it, it won't be able to be used "efficaciously"; for example, in a sentence. As a consequence,someone who has properly formed concept should be able to use that word notably better than someone who hasn't. This provides us with a straightforward basis to test Rand's theory, so here's a couple of quick thought experiments as to how this could be done.
1) Take two young children of similar intelligence who are at an articulate age but have not observed some particular object - let's say, for example, a microscope. Then take two microscopes of identical make and model. Show the first child the first microscope, then remove it and show her the second microscope. Tell her that these are represented by the word, "microscope." Having seen two "units", the child should, according to Rand's theory, have now formed a concept and be able to use the word in a sentence to some degree. Now, do the same with the second child, except this time instead of showing her the second microscope, show her the first one again. Thus, she will have only observed one unit, not the two or more required by Rand's theory, and thus should have not correctly formed a concept. Thus her verbal performance when tested should be markedly worse than the first child.
2) Another possibility is to imagine say, a child in remote village. Let's say the village has one piano, which the child is taught to play over the years, eventually becoming very proficient, coming to understand how the parts work, how to fix it etc. Let's compare this to another child who pauses at the window of a piano shop in a busy city, where they see two different pianos briefly, before their parents hurry them on. Now, according to Rand's theory, the second child will have a properly formed concept of "piano", whereas the first won't, and should be more "cognitively efficacious" regarding pianos than the first child is.
Of course the thought of either such result actually occurring seems absurd, which seems to indicate that Rand's theory is equally absurd. Naturally this could be down to the design of my thought experiments, in which case dissenting Objectivists are welcome to make better suggestions.