I simply don't buy that in any one instance, the full course can kill all bacteria (with the help of the immune system) while the half course couldn't (with the help of the immune system)
Who said it couldn't? The problem arises when half a course or irregular administration of antibiotics results in organisms sticking around longer and making resistance more likely.
This whole thing is just a numbers game. In fact, there seems to be a sweet spot where resistance is more likely to happen -- it isn't as though whole course/half course is the real issue.
The real issue is how effective the antibiotic course is in eradicating organisms. If it is highly effective then you can't get resistance because all the organisms are gone in that particular body. If they aren't very effective, then resistance won't develop very easily because it won't matter much if the organisms are affected by the antibiotic or not. Any bacteria that develop resistance may be able to stick around longer than the others, but since there are always many other factors at play to kill them (immune system, bacteriophages, other bacteria, etc.) the resistant strains are no more likely to make than the non-resistant strains.
In scenario number one (which does happen) there are no bacteria left to be able to produce variability. In scenario number two there are plenty of bacteria to produce plenty of variability but it doesn't matter much if resistance develops.
Resistance does develop in scenario number three -- where there is a strong selection pressure and there are enough bacteria left around to produce lots of variable strains in which a resistant strain is one possibility. That can happen most easily if you have a highly sensitive bacterium and apply only a relatively mild pressure that works but does not eradicate them all (only taking part of a course of antibiotics) or if you have a bacterium that is not going to be completely eradicated by the antibiotic but is still placed under significant pressure. Resistance is likely to develop only if there are still quite a few bacteria left around to produce a significant degree of variability. If there are only a few bacteria floating around and they are kept in check by the immune system at a very low number, then they are simply not going to produce many copies, so there is less chance of new mutations resulting in resistant strains. With this scenario, the limiting factor is the host's life. The number of possible mutations is a product of the number of organisms and time. Eventually resistance should develop in every case where bacteria are not eradicated completely except for the fact that hosts don't live forever. The bacteria in them usually die with the host (many can survive through other means, though). If there are many bacteria left around and the selection pressure is high, then resistance is going to develop. The speed of development depends on the number of bacteria produced in general. It's jsut a numbers game.
So, I'm afraid it just isn't as simple as half-course/whole-course and you've got an answer. It depends on what type of organism you speak of.
You aren't dying when you're doing the half course. You get better, because the immune system has the pathogens under control. Cleaning up is just aftermath. Of course you're still somewhat vulnerable to relapse, which may well be the true reason why you're supposed to take the full couse.
Dying, not dying doesn't matter for the most part. That is a host response and the bacteria don't care. The important matter is the number of organisms left over by a half course vs a full course. It depends on what antibiotic, what organism, and the hosts immune system. There are simply too many factors at play that no one can predict in all cases what will happen.
Besides if half a course is just as effective as a full course, then half a course should produce precisely the same pressure as a full course and exactly the same amount of resistance.
But regardless of me being right about this or not, if you add in the plasmids to the calculation, and the probiotic flora (which is in some cases itself acting pathogenic) in and on the body that will also more than likely develop resistance, which it can share via the plasmids, even over the species barrier. And this probiotic flora cannot be realistically killed because it is almost constantly replenished externally. This alone speaks even more strongly against your theory.
First of all this isn't my theory. This is all just basic evolutionary theory.
As to the other flora, yes, that is a huge problem. The existence of plasmids and the natureal flora in our bodies is one of the reasons why we are going to lose the battle against resistance. It really is just a matter of time. The solution is not half-courses of antibiotics, however. That won't solve anything. Half courses are still going to produce resistance over time.
We're going to lose. That's all she wrote. It's basic biology.
Selection pressure creates selection. You are assuming that "too much selection pressure causes extinction" but the reality of multiple-antibiotics-resistant strains of bacteria proves you wrong. These bacteria didn't occur because someone took an insufficient quantity of antibiotics, they've always been there. The antibiotics - of a full course or a half course - simply removed all the others, and the resistant strains replaced their Lebensraum with ease.
Multi-resistant strains of bacteria prove nothing of the sort. What they prove is that there are multi-resistant strains of bacteria. There are many resons why we have all these different strains, but it is not because drug companies and doctor's promote taking full courses of antibiotics to treat disease and yell at people if they only take half-courses. It is the antibiotics themselves that promote resistance whatever the course.
We do not know if these multi-resistant strains have always been there or if they developed by means of new mutations. Both are possibilities. It is very likely that they were always around at a low level and not particularly selected for. But there is no reason to expect that they would continue to harbor resistances if the presence of resistance did not provide a selective advantage. Other things do kill bacteria after all. Those resistant strains could just as easily be killed off by some weird contingency.