Sorry, that's a fallacy of "there's still a chance, right?" It's all probability, but probability isn't a one-color system of grey, it's many shades of grey, and you can be "more white" and "more black" (though never pure white or pure black; I'll explain this in the shortest way possible by saying that 0 and 1 are not probabilities, just as plus and minus infinity are not integers). For some reason people think that this ideal "knowledge", for example 0.0% chance of winning the lottery, somehow means you have to cast aside the belief, whereas if the chance were 1/googolplex, you're allowed to keep it.
Failure to distinguish this is the real essence of the "bad subjective" in which the first post decries. If a person is going to ignore your argument that says the likelihood is 1/googolplex, why don't they just as well ignore some argument which claims to have 0 likelihood? And conversely, if you're going to ignore an argument that's 99% certain, you might as well ignore an argument that's 100%.
Though your terminology of "pre-knowledge" isn't entirely incorrect... Probabilities exist in the mind; I can flip a coin and assign 50% chance it lands on heads, but reality isn't uncertain and has already decided, and if I have the option of repeated experimentation, I can calibrate my probabilities to make my map better reflect the territory by seeing which side of the coin comes up. (Ignore quantum stuff.)
Really? I'd like to see that proof. But don't spend your time, because what I'm really at odds is the redefinition of objective and subjective, which lead to arguments over definitions. What great light in the sky says your values are objective, and good for you? Death could be considered an objective fact by most, I think, but I don't think it would be a good value to prefer it on the grounds that it's objective, and I intend to live forever (provided there's enough fun) anyway.
Intrinsic is a part of the mind projection fallacy, where you project your map onto the territory, your mind's perception of reality onto reality itself. Your mind is not separate from reality, but you are not a god, and can't discern reality as it Really Is, sitting in your living room with your blinds shut. We've only known about quarks for so long, you know, and that knowledge came around through experimentation. But so did atoms, so just because you experimentally verified something doesn't mean the territory (reality) actually works that way, just that your map of the territory at that level works that way. We know that the universe doesn't change its calculations for things at different scales; it's all quarks (as far as we can tell). But we have to use less-precise-but-faster-to-compute methods because our computing power can't handle the everything-is-quarks map yet. I agree that intrinsic thinking ought to be avoided.
"Subjective" in this case is the on-the-whim decisions which Rand discouraged and so do I. But people really make many of their decisions on this sort of basis very frequently (right now I just feel like posting), and then they rationalize why it was rational later (I'm building my writing skills, laying down my beliefs so I can see them more clearly, challenging what I believe to be false statements, etc.). This isn't necessarily a bad thing; if you're faced with a situation in which you must either run into a burning orphanage to save someone or give in to your fear of death and run away, you're not going to have a lot of time to weigh the reasons for or against one or the other, and if you sit down and think about the problem you can find many. You'll start what-if-ing, coming up with different solutions for different scenarios (go in if it's a family member you love, run away if firefighters are already there and the heat is too intense for you), but I believing finding general solutions are of much greater use. This is my objection to people's objection against thinking about hypothetical situations, especially broad ones. Argue it out now and come to as rational a solution as you can, getting as least specific as you can, then in the event that you're faced with something like it you'll know what to do (if you still value rationality, anyway).
"Objective" in this case is still very subjective, but I'll grant that "objective" in this case is really just a redefinition so more distinct categories can be made. I think it might help the Objectivist cause more if they were labelled Garf, Pappit, and Morpid, along with as clear-as-possible definitions of each that don't just say "Pappit = subjective". An ideal definition would be to say "these configurations of quarks represent the human brain acting Garfily, and are to be avoided". But that's not likely to happen. It's important (I've found) when talking to people about selfishness that it's not at all used in the classic sense of the word which society still holds, the one which carries a negative connotation, or the one which goes along the lines of "caring about your own life regardless of anyone else's". I think Objectivism would have less resistance to the ideas if new words were used instead of hijacking meanings from other established words. I can say "the fox is asleep" and really mean "my head itched, and a scan says my brain is receiving itch signals from the scalp-nerves, and continual scanning will show my brain telling my hand to scratch it", but I'm going to find it hard conveying that information. Defining selfishness as "concern with one's own interests" (Virtue of Selfishness) is okay, but then you have to also define what interests mean, and what constitutes concern, and what those things are, and so forth. Then you start bringing "rational self-interest" in it, without any math to back you up, and you're done. (I'm firmly convinced that the art of rationality lies in the studying of probability and decision theories.) Once you start using abnormal meanings for common terminology you've got to be very, very explicit. The nice thing about math is that it is explicit.
Anyway, "objective" is subjective ("in the mind", which seems to me a standard definition, if somewhat poorly defined) in this case because you're taking your past experiences of eating bananas as evidence, along with the people you know's past experiences. I can give the same argument for why I should eat dark chocolate. Except try feeding chocolate to a cat (not seriously, please). To me, for something to be objective it seems like it should apply regardless of the situation. For it to be "subjectively objective", which is as close to objectivity I believe humans as they are now are to get, it has to exist in the most accurate map we have of the territory. Back in Newton's times, we would say that his equations were objective because our maps which used them reflected the territory better than any others, in that they produced more accurate results and so forth. Then later on Einstein showed up, and helped us draw an even more accurate map, showing that it has really been Relativity governing things all along, even in Newton's times, and Newton's equations were just special cases of Einstein's.
Now as I'm at the end of my post (more observant users will notice it's my first; this felt like a good place to come in instead of the traditional method of making a greeting topic, saying nice flattering things all the time, then every once in a while raising a few questions or concerns), I want to reveal a little of my personality along with an attempt to prevent ad hominem attacks. I'm not here to blindly attack everything to do with Objectivism or its followers or all of its beliefs (you won't find me dissing laissez-faire capitalism, for example, or advocating communism). I would like some sort of intelligent response though instead of one consisting of just pointing me to literature, though, as I've had that in the past... And while the real "enemy" is the irrational leaders who threaten to blow the world up, or the irresponsible, over-arrogant engineers who dream nanotech or strong-AI without considering all the precautions, I do think myself as less wrong on a lot of things; but I strive to become less wrong, and I'm not going to stop when I really think I have the answers. I'm not going to accept anything with absolute certainty (even this sentence), because I would never be able to undo it. I've found I don't need absolute certainty; 90%, 95%, 99%, 99.99...9% (with 3^^^3 9's, a very arrogant percentage) work for me.