First let me just give you a philosophical overview as to why the question is disingenuous and you are barking up the wrong tree. You ask:
Was Richard Leakey correct, did they toss out the fossil, or the theories on early man?
No, he was not. Of course, that is because he was taken out of context in that selective quote (a favorite creationist tactic). I'll lecture later on why he said it, but later in the same article, he also admits that he, or someone else could have made a mistake, and they needed to figure it out.
Were the first 4 (2.6. Million) dates thought to be accurate at the time of their original publication?
Not exactly, because again, this is creationist propaganda. If you actually read the cited papers, they aren't actually dating the skull, but other material found in the same strata as the skull, and in locations they thought were similar. Also, note the dates of all those publications. Methodologies have changed since the 1970s. Furthermore, we understand the geology of the area better, which has cleared up a lot of confusing results. Note that the skull was discovered in 1972, yet the creationist page uses papers from 1970.
Were they later "revised" to the more recent 1.8 million years ago timeframe?
Because that's what the evidence indicated they should do. Not because of the old and inaccurate radiometric results, but because they figured out their mistake... The ability for science to be a self correcting mechanism isn't a weakness or a fault, that is the strength. You seem to indicate that this is an unacceptable practice, yet I am sure the computer you are using isn't from the 1970s.
Is this evidence that the fossil record trumps all other dating methods (i.e. if there is disagreement, the fossil record is not revised, it is the other dating methods)?
No. The fossil record really isn't required to have evidence for evolution. However, what the fossil record does is give us a more detailed understanding of evolution. Of course, there aren't Post-It(TM) notes attached to fossils describing all the circumstances of fossilization, what exact species it is, or exactly how it fits in to the evidence trail. That's what the scientists need to figure out.
Okay, now to some more detailed rebuttals, and again, I am sorry if this seems like a lecture, but all to often, creationists will try to use one single misunderstanding to attempt to "prove" their point, while totally ignoring the mountain of evidence that makes their stance the height of ignorance and foolishness. They trot out the same lame arguments time and again, and never listen to the actual replies.
As to Leakey's quote, as a scientist, he understood that when you find an earth-shattering discovery that goes against or disagrees with what we know, there are several possibilities, two of the most commonly thought are:
Scientists want and desire to make the first type of discovery. It's what puts them into the journals and history books. What usually happens is the second one. In the case of KNM-ER 1470, it was indeed the latter, but with a touch of the former. In this paper, Leakey admits to his mistake about the initial date and the KBS Tuff:
At present, analysis of samples collected for dating from the KBS Tuff in area 131 has proved inconclusive because of the apparent alteration of the sanidine felspars. This was not seen in the 105/108 samples from the same horizon which provided the date of 2.61 m.y. and there is no reason to suspect the validity of that date (personal communication from J. A. Miller).
However, as a result of all this investigation, he did figure out that he may be on to a new species in the genus Homo. Specifically Homo rudolfensis. Now that we have figured out that there is a new member of the genus Homo, the fact that creationists keep using this 40 year old argument, goes back to the fact that they don't listen... As the archeology info site states (emphasis mine):
The specimen was originally thought to be around 2.9 myr old, due to an inaccurate dating of 2.6 myr for the KBS volcanic tuff located above it. This inaccuracy was caused by contamination of older material, and the tuff is now know to be much younger. The specimen is now thought to date to approximately 1.8 myr (Leakey et al. may have been more willing to attribute the specimen to habilis had they known the real antiquity of the specimen from the beginning). Though this date is now generally accepted for the specimen, the geologists who orignally dated the KBS tuff continue to argue for a later date for the specimen. While they admit the dating of the volcanic tuff was inaccurate,
This page doesn't quote anything after 2000, and the 2001 paper with Leakey et al does figure out the problem with the Tuff.
One other problem that many people ignorant of human evolution fail to understand is that it's not a linear path. As a previous link mentions, several hominid species co-existed. Some became extinct, and some didn't. This is where Leakey actually made a great new discovery. Again, it's all part of the self correcting mechanism of science. Something was off, and at first they didn't know. Further investigation and scientific methodologies actually gave them clues, and then they had to revise what they thought they knew. This is the accepted scientific method, not blindly accepting the first thought that may come to one's mind. A paper that details a better understanding of the family tree was published by Bernard Wood (one of the men that help assemble the skull with Leakey:
Discussing the radiometric dating as it was understood in the 1970s compared to the 1990s and beyond, Dr Groves of Tufts university relays some information on KNM-ER 1470 that may be relevant:
White et al wrote of Aramis: "All hominid specimens were surface finds ..." This does not mean that they were rolled in from elsewhere; in fact, their condition indicates otherwise (there is a whole field called Taphonomy which is devoted to the study of how fossils got to be where they are). The date was also, at one point, queried by Kappelman and Fleagle (1995, Nature, 376:558-559), and satisfactorily answered by Wolde Gabriel et al. in the same edition of Nature.
This is an account of the history of the growth of understanding of the dating of the deposits; it is not some kind of admission of circular reasoning, of making the 40Ar/39Ar dates fit. They then go on to explain in some detail why Kappelman et al. misunderstood the argument about the dates.
[As an aside, there has been much speculation, informed and uninformed, about the reasons for the delay in publishing further information about Ardipithecus. White presented a paper at a congress in South Africa in late June/early July, 1998, attended by about 750 palaeoanthropologists and human biologists. The paper was illustrated by slides of the site and some of the fossil material. The site is flat, stony and arid; the fossils are scattered over the landscape, friable in the extreme, and difficult to collect, let alone to preserve. It is true that remains of at least 50 specimens ascribed to A.ramidus have been collected, but these will take years to preserve, fit together, study and describe.]
For Lubenow, this seems to be an attempt to make the radiometric dates fit the faunal analysis, which reminds him of the polemic over the date of the famous skull KNM-ER 1470, from Lake Turkana in Kenya. There was, he claims, such disagreement between radiometric and other dating techniques that it was finally dated by "biochronological comparisons", in this case the stages of evolution of pigs in East Africa.
Not true, actually. What is true is that the palaeontologists argued for some years for restudy of the K/Ar dating techniques, because their faunal analyses suggested that the date of the KBS Tuff, below which ER 1470 was found, was anomalous. Ian McDougall’s group finally managed to date material from which all contaminants had been eliminated, and showed that the palaeontologists were correct.
[Incidentally, the significance of these "stages of evolution of pigs" is passed over in silence by Lubenow. Surely he is not making a tacit admission that pigs did evolve! In creationism, neither humans nor pigs should evolve; and yet, there they are - several lineages of them, abundantly represented - in sites like Omo/Shungura, whose stratigraphy and dating are entirely uncontroversial, evolving away and eventually either becoming extinct or ending up as the modern Bushpig, Warthog and Giant Forest Hog. Elephants, incidentally, did the same thing, and their remains are likewise sufficiently numerous that their evolution can be tracked in great detail between about 4 and 1 million Ma.]
Again, this Taphonomy is a delicate science that may require decades to fully understand how something happened. And initial mistakes are remembered, while the correct answer seems to go ignored. Of course, this is nothing new from creationists.
If you need to get a Christian perspective, this individual takes a biblical approach, and still has to conclude that radiometric dating works and is accurate. A telling bit in that paper states:
Other creationists have focused on instances in which radiometric dating seems to yield incorrect results. In most instances, these efforts are flawed because the authors have misunderstood or misrepresented the data they attempt to analyze
Which seems to be the entire premise of the KNM-ER 1470 discussion. Nothing but misunderstanding and misrepresentation.