First off, I want to make sure you realize that theory has a different meaning in science than in general parlance. A scientific theory is “a plausible or scientifically acceptable general principle or body of principles offered to explain phenomena” (as m-w.com puts it). A chemistry textbook I have here goes even further, calling a theory “an encompassing idea that provides a full explanation for a natural phenomenon”. Theories aren’t nearly as tentative a thing in science as most people think (the atomic theory of matter is an example of a theory, but it’s pretty much definitely how things really work).
Laws are usually over-arching, invariant principles, such as the second law of thermodynamics (for any spontaneous process, the total entropy of the universe must increase). Laws aren’t necessarily more true than theories (the ideal gas law works, but is really only true for an imaginary ideal gas, for example), they’re just broader.
Ok, now with that out of the way, let’s see how I can do on immunology. Note: my knowledge of antibodies is much, much, much stronger than my knowledge of immunology within an organism–just so you know what to put the most validity in… but I’m using an Immunology textbook to supplement my memory, so this SHOULD be right.
Also note that I’m oversimplifying–we also have T cells and T cell receptors, which adds to our ability to find and clear antigens… but they just make the numbers bigger, the general principles are still the same
During B cell (the cells that produce antibodies) maturation, the genes for antibodies are shuffled to form the final antibody gene for that cell. Each cell produces its own combination, so each cell is different than its “sisters” (theoretically the same combination COULD come up more than once, and probably does). This shuffling alone can account for more than 10^8 possible specificities. Somatic hypermutation (a million-fold higher mutation rate in certain regions of antibody genes versus normal genes) increases this to on the order of 10^11 specificities. Note: this number is reduced during maturation, because antibodies that recognize things normally in the body need to be eliminated. I’m mostly ignoring that problem in this discussion.
So, after your immune system has “grown up”, you have all these different B cells in your system, producing different antibodies. When a B cell is presented with an antigen that it can bind, it reproduces, producing many copies of that same antibody (that’s the clonal selection).
Why isn’t the human body invincible? There are many factors.
First, while there are 10^11 different possibilities, that still doesn’t necessarily cover every possible antigen. For something as large as a virus, though, there will almost definitely be a B cell with an antibody capable of binding SOMETHING on the virus, so let’s move on (the “lack of coverage” comes in more in experimental practice, when you’re trying to, say, make an antibody to TNT–there isn’t necessarily an antibody anywhere within your experimental mouse that has good TNT-binding ability).
Next, the B cell with the specificity required has to encounter the antigen in time. That’s why immunizations work–you increase the number of B cells with the “good” antibody, so the chances of it finding the antigen in time are much, much higher.
So, how do you get diseases? The disease takes hold before the B cells that could fight it off are able to find it. HIV is a special case. Since it attacks the immune system directly, it makes it even less likely that it’ll be found in time (there’s nothing left to find it, essentially–again, this is an over-simplification, but you get the idea). Also, the most potentially immunogenic portions of HIV have extremely high mutation rates (only certain parts are necessary, and these necessary parts are burried in mutable “junk”, essentially). So, say your body fights off a certain batch of HIV… it’s still susceptible to HIV, even HIV that’s not many generations removed from the viruses it already fought off. Another issue is that HIV (like many other viruses) “cloaks” itself in a membrane made from the cell membranes of the person it’s infecting. Your B cells can’t always see it as anything foreign at all (except I’m fairly certain it displays a few proteins on its surface that are still foreign, so it’s POSSIBLE to recognize it–just more difficult than it would be if the entire outside coat were foreign).
As for injecting monoclonal antibodies into someone with a disease… well, that works to clear out something that isn’t increasing (such as snake venom–that’s exactly what antivenom consists of), but you need the person with the disease to keep producing antibodies to keep fighting off the disease (otherwise one virus can get away and start the disease all over again). Chances are the person would clear out those injected antibodies with antibodies of its own too soon for them to be effective (taking into account blood group doesn’t matter–the antibodies themselves are perceived as antigens, because they’re still foreign).
It IS possible that we can inject something that will help find those B cells with the “good” antibodies, though. That’s what immunizations consist of. The reason some immunizations can be effective even when someone already has a disease is that they help increase the “good” B cell count without the B cells having to find the “true” antigens. With increased knowledge of how antigens work, it’s theoretically possible that all diseases can be destroyed, yes… we just aren’t sure how to do it in all cases yet.