Dev, you're using the common yet insidious definition of atheism and agnosticism. The atheist from your definition is a straw-man that is easily attacked and brought down. In reality the vast majority of people who self-identify as atheists fit your definition of agnostic. Let's first clarify how I, and most of the self-identified atheists, would define the term and then look at why I label the definition you're using as insidious.

The common definition of these terms assumes a religiosity spectrum with atheism and theism diametrically opposed and with agnosticism as a middle ground. Instead, let's define atheism/theism and agnosticism/gnosticism as two binaries that answer completely separate questions. If you are asked whether or not you believed in God you would use the first binary, and if asked whether or not you are certain of God's existence/nonexistence you would use the second binary.

In other words, a theist is someone who believes in God. If you are asked the question, "Do you believe in God," unless you can answer with an affirmation, you are an atheist. If you are certain of your answer to the previous question you are a gnostic. Otherwise, you are an agnostic. So you can either be a gnostic theist, gnostic atheist, agnostic theist, or an agnostic atheist. At this point it should be pointed out that those who express gnosticism are almost exclusively theists.

Now that my definition of these terms is clear, let's address why my definition is superior to yours. First of all, it is etymologically correct. If you break the word atheism apart it literally means without theism. Furthermore, if you look at the word gnosticism it means pertaining to knowledge. Theism and gnosticism inherently address fundamentally different things, so your definition seems incorrect to use them as part of the same spectrum. Secondly, my definition is more consistent with the way things actually are since my definition of atheist includes both gnostic atheists and agnostic atheists, whereas your definition of atheist is limited to just gnostic atheists. As I've already pointed out, the gnostic atheist is a beast not really found in nature. Thirdly, my definition provides more information than your definition as your definition offers a triad of options about your knowledge/faith whereas my definition offers a tetrad of options. Finally, my definition does not allow the straw-man fallacy to be used as I show below.

Since the etymology of the terms has been established, it should be clear that the theists have hijacked the word "atheist" for their own purposes. Because, as you correctly identified, the certainty of the nonexistence of God is as illogical as the certainty of his existence, your definition allows the theist to put atheism on a philosophical foundation as questionable as theism itself. The almost nonexistent (gnostic) atheist is easily brought down, and then the claim is made that (both gnostic and agnostic) atheism is philosophically foolish. This is obviously a pernicious abuse of language, and should not be allowed to stand. This is why I have long advocated for the use of my definition, and why I call your definition insidious.

So in light of what we've discussed, it seems to be the gnostics that are the issue according to you, not the atheist or the theist. I point out again, however, that gnosticism is implicitly linked with faith. You can't definitively say anything about the supernatural without a rigid belief in something. All atheists say is that we have no evidence for God. So if gnosticism is the problem, the theists are exponentially worse than the atheists.

But what about the agnostics? Are the agnostic atheists as bad as the agnostic theists? Even here, I would argue that the theists are on epistemically shakier grounds than atheists. I could point out things like burden of proof, or use science in conjunction with Occam's Razor to make this point, however that is a discussion for a different thread.

But ultimately the point to this post is that atheism and theism are different. They make entirely different points using entirely different arguments. Equating the two is specious.