Labels have a wonderful, inherit quality of being both necessary and useless at the same time. They can unite, divide, separate, or confuse. To this end, when there is a conflict involving labels, it's important to flush out the discrepancy. Some quick hits:
Atheism is not a positive position; it is a label for those who reject a theistic position. Without theism, there is no atheism.
Atheism is not a descriptive label of a persons affirmed beliefs. There are a nearly innumerable number of labels that atheists use to describe their actual beliefs, such as secular humanists, rational skeptics, pearlists, objectivists, etc.
Atheists are not necessarily rational people. Atheism is simply the rejection of theistic claims. Raelians, Scientologists, certain Buddhists, certain Satanists could all be considered atheists. Some atheists are even believers in the Paranormal, followers of alternative medicines, conspiracy theorists, anti-vaxers, utopian communists/libertarians, or utilitarians.
Back to the topic at hand, Gnosticism, at its root, deals with a claim of absolute knowledge. Agnosticism is the claim that you can't claim absolute knowledge. As you may or may not agree, outside of certain logical absolutes, any claim of or claim about absolute knowledge is useless: brain in a vat, solipsists, etc. Agnosticism and atheism are
not mutually exclusive terms. They are, however, commonly presented as such. When agnosticism is presented as this middle ground between theism and atheism, it has the side effect of straw-manning all atheists. It redefines the negative claim (rejection of a positive theistic claim) into a positive claim that there is absolutely no god, a fairly indefensible position. The vast majority of atheists that I have talked, discussed, or interacted with(noted obvious sample bias) make no claims about the existence of a god or gods. They simply have rejected as insufficient any claims of a deity that have been presented to them. This is why agnosticism is a description of belief rather than a position of belief. There already exists a framework as can be seen here:
http://freethinker.co.uk/2009/09/25/8419/. I do have some qualms with the link, so it's more of a guide than a canon. Most people fall into the camps of gnostic theism or agnostic atheism. Gnostic atheists could include atheistic religions or emotional atheists. Agnostic theists would be mostly deists.
I do accept these labels as valid, but I don't think they are sound. They are still very useful for everyday conversations about belief and deal with a problem of agnosticism quite well. I am only discussing this caveat in response to your post about the scientific method and agnosticism:
Agnosticism is not a preference but a method of reasoning which states that one should not accept as certain any conclusion that is not demonstrable. It is the basic of the scientific method and is the only reasonable means of regarding untestable propositions.
I would have to wholeheartedly disagree. Agnosticism,especially in the form presented by those who proclaim that position and reject atheism, is inherently unscientific. When one is presented with a claim, he could either accept the claim or reject the claim as insufficient. In every day life, most claims are either accepted or rejected without much scrutiny. This usually depends on social relationships, authority, or the importance of the claim. If I were to tell you that I don't have any dogs, you would most likely accept that as true. I don't have a reason to lie to you, and the claim is relatively insignificant to you. As the importance of the claim increases, the necessary scrutiny for that claim to be accepted increases. If you were over my house and were bitten by a rabid dog, my claim about not owning a dog has suddenly become very important. Even though the claim is about not owning something, there is still a positive testable claim. It would be your claim, however tenuous, that I do own a dog. You could check my house for dog food, dog bowls, collars, leashes, or fur. You could contact the town or local vets and see if I have any dogs registered. You could check my credit card statement. You could talk to my neighbors. If, after all this searching and investigation, there was no evidence that I own a dog, would it be reasonable for you to claim that I own a dog? Would it even be reasonable to say that you can't "know" whether or not I own a dog? If a rabid dog, matching the description you gave, was found to be owned by someone down the street, would it still be reasonable to claim that I own a dog? Would it be reasonable to claim that you can't "know" whether or not I own a dog?
Let's assume your search was fruitful. You found a bag of dog food in my garage. You found a dog collar with the name 'Spike'. You found a TV that plays an endless loop of All Dogs Go to Heaven, The Lady and the Tramp, and Homeward Bound. The local vet has a record of me. The town says that there is a dog registered at my address. A receipt shows that I recently spent 35 dollars at Scrub'a'Pup. Several neighbors claim to have seen me walking a dog that fits the description you gave. Yet, you still haven't found a dog. Would it be reasonable for you to assume that I don't have a dog? Would it be reasonable to claim that you can't know whether or not I own a dog?
Science has never been about absolute knowledge. It never has, and it never will. If man ever reaches the point where we know everything, then science simply stops. Science has also never been about proof; it is about evidence. Science can not prove the sun will rise tomorrow. Science can show evidence of all the previous occasions the sun has risen. It can show that most stars die around age x, and the sun is at age y. It can show that there has been no irregular activity from the sun indicating its collapse. It can show piles and piles of evidence against an implosion of the sun tomorrow. It can still never prove that it won't happen. There exists all of this evidence that leads to the conclusion that the sun won't implode tomorrow. There is no evidence suggesting that it will. What is the reasonable position for someone to take, since there is no proof either way? Is it to say that we can not know whether or not the sun will implode tomorrow? Or, is it more reasonable to reject the claim that the sun will implode tomorrow as insufficient?
I reject the claim that the sun will implode tomorrow. If evidence is presented indicating the star's collapse, then I will reevaluate my position. Until then, I am justified in assuming that the sun will rise tomorrow. I also reject the claim that leprechauns exist. If evidence is presented indicating the existence of leprechauns, then I will reevaluate my position. I also reject the claim that I am Julius Caesar reincarnated. If evidence is presented that indicates that I am perhaps Julius Caesar reincarnated, then I will reevaluate my position. It is not scientific to say that because we can not ever "know" anything, we can't reject claims. Science frequently rejects claims as unfounded, insufficient, or unsupported. Science throws out previously held beliefs when they contradict the evidence.
There is no current study on the existence of leprechauns because there is no evidence to test. No scientist has ever approached me to examine the possibility that I am Julius Caesar reincarnated. Why? There is no evidence to suggest that it is true. There are scientists, however, smashing atoms in Switzerland looking for the Higgs-Boson. There is an observable phenomenon and a proposed hypothesis. The scientific position on the Higgs-Boson is that it does not exist. Existence is a claim, whether we are discussing the Higgs-Boson, leprechauns, or a deity. Until there is sufficient evidence of existence, it does not exist. It is perfectly reasonable and justified to claim that no god exists, until there is sufficient evidence of its existence.