I can't imagine scientists think "as far as we know, we haven't managed to decompose that stuff yet and in so far as we don't even know at all what it is, let's assert it's an element until detractors crop up" that sounds wrong. I just seek to gain an insight.
Yup, that IS the thinking (or what the thinking "should be" anyways). Scientists do experiments. The results of those experiments either support or contradict existing theories. When they contradict existing theories, then new theories have to arise that cover both the old and the new observations. Then these theories are tested against future experiments until some experiment comes along to challenge it, and so on and so on, etc. In other words, science isn't a static thing...it changes with time. Our "view of the world" changes with time along with it. The science of the 1700s is not the science of today. "The truth" from science in the 1700s is not the same "truth" of today (unless you consider that the truth of science was never meant to be absolute, which it isn't, in which case you would say that the truths of the past have merely been modified several times over to account for newer experiments).
We tend to overlook this even now (but Asimov doesn't, hence the value of his books!) when we describe science because it makes us feel better to believe that what we understand at present is "the absolute truth". For example, we claim that Democritus discovered the atom in ancient Greece, and that Dalton "proved it," and from then on the idea was accepted as "a truth" by scientists in happy harmony. It was simply not so. In fact, it was not until the early 1900s, when experiments started to demonstrate the structure (and substructure) of the individual atoms making up matter, that thinking about atoms became mainstream. There were MANY scientists in the late 1800's that fought bitterly against atomic theory, and some refused to accept it outright even into the 1900s. This came from the prominent discoveries of the past, which these sceintists were having a hard time letting go of. One such example was that of electricity. Benjamin Franklin imagined electricity as a continuous fluid, and not as an assemblage of discrete particles. Fluid theories were popular and existed in other forms such as the "caloric" for energy and theories at the time had to be explained in terms of a fluid being continuous to even be generally accepted...whether or not the experimenter believed that matter was continuous or not. This was part of the reason that Arrhenius graduate thesis in 1884, on the existence of "ions" in solution to explain electrical conductivity, was not only opposed at the time...he barely graduated with it (with the lowest possible distinction) despite the fact that it was argued so well that nobody could really oppose it. Now it is accepted (and Arrhenius got the nobel prize in 1903 for his ionic theory once some of the evidence that matter was particulate in nature had come streaming in), but at the time this line of thinking was frowned upon.
Perhaps one day our modern view of atoms will be gone (!) and we will be describing interactions between some new type of thing smaller than an atom (or perhaps, larger!) that someone will describe that can make sense out of quantum mechanics (but even to do that you have to overcome the modern thinking that quantum mechanics is "supposed" to not make sense by its nature...a philosophy put forth by Bohr and company largely, in my view, in an attempt to protect their theory from future theories and advancements). To do anything like that, you have to wait on the results of experiments. At present time, physicists are waiting for data from the Large Hadron Collider to see if the data that comes out supports our modern views of quarks and whatnot or if it opens the door to some new understanding that we can't really anticipate at present (actually we are merely waiting to see if the thing will ever even work...but it sounds more exciting to say it the other way
I will wait till they find something right then.
Even if our modern understanding of science is incomplete, taken as a whole you can use it to predict things. There is its value. Any branch of knowledge has its value in this sense; the ones that are the most accurate in their prediction of the results are the ones we depend on the most when we go about influencing our lives. Even wrong theories like the Phlogiston theory that are thrown out to make way for new ones were useful at the time because they helped thinkers predict and understand what it was they were trying to do at least in some cases. If you are waiting for the "absolute truth" to come along so that you never have to refine your views, you will be waiting (in complete ignorance) for a long time. We don't have it, and I don't think we ever will (but then again, this is just my opinion...).