In my view, John was developing Jesus' relationship to God in primarily functional categories in the Prologue, and in the rest of the Gospel. I have no problem with either a definite or an indefinite rendering of John 1:1c, as both "the Word was God" and "the Word was a god" are grammatically possible and theologically and contextually appropriate when interpreted properly. But for most orthodox Christians this verse is perhaps the most sacred of all theological cows, which must either state that Jesus is "God" definitely (i.e. identifying him as the one God of the Bible), or that he is God "qualitatively," which is presuppositionally nuanced to mean that he fully shares the 'divine nature' with the Father within a trinitarian Godhead.
As I read the various arguments by people who are referenced to support taking QEOS "qualitatively" at John 1:1c -- i.e. Paul Dixon , P.B. Harner , Don Hartley , etc -- it became very clear that these folks are theologians, not professional linguists. If you compare a good secular study in linguistics and grammar with these studies, you'll probably be struck by how qualitatively different the religiously motivated thesis is.
Unfortunately, it's pretty clear that the "qualitative" thesis has never been properly vetted. Rather, it was embraced with alacrity because it provided exactly what the doctor ordered, namely, a solution to problems that began lurking in the shadows as people came to realize that the previous answer probably wasn't the right answer after all. What was the previous answer? It was, as so many had argued prior to the 1970s and some continue to argue today, that QEOS is definite at John 1:1c, per Colwell's Rule, which was articulated by E.C. Colwell in 1933 . Colwell's Rule as (mis)applied to John 1:1c was also embraced uncritically and with great alacrity for many years, but then folks began to perceive problems.
What were the problems? Firstly, Colwell's rule states that "Definite predicate nouns which precede the verb usually lack the article,"  yet most theologians, including Colwell himself (it seems), assumed that the converse were true, i.e. that anarthrous predicate nominatives that precede the verb are usually definite, which is obviously incorrect. If you review part 2 of this series, you'll see that there are many anarthrous predicate nouns that appear before the verb in the Greek NT, which are clearly not definite. Secondly, theologians came to believe that if QEOS were definite at John 1:1c, then this would result in Sabellianism. This isn't necessarily correct, but it is a commonly accepted presupposition. So these problems -- one legitimate and one assumed -- necessitated a 'new solution' to the problem of John 1:1c.
Let's start with the last item, i.e. what motivates some to embrace and promote the notion that QEOS at John 1:1c is "qualitative"? Paul Dixon gives the game away on page 2 of his DTS Masters thesis, The Significance of the Anarthrous Predicate Nominative in John, to wit:
"The importance of this thesis is clearly seen in the above example (John 1:1) where the doctrines of the deity of Christ and the Trinity are at stake. For, if the Word was 'a god,' then by implication there are other gods of which Jesus is one. On the other hand, if [QEOS] is just as definite as the articular construction following the verb because, 'the dropping of the article…is simply a matter of word order,' then the doctrine of the Trinity is denied."
"An indefinite noun is a noun which stresses neither definiteness, nor qualitativeness, but membership in a class of which there are other members. Technically, any noun which is not definite is indefinite. For expediency, however, we exclude qualitative nouns from the class of indefinite nouns." (ibid, p. 9)
Interestingly, definite nouns can also be used to stress 'nature,' and, once again, the "qualitativeness" that we infer depends on the noun's definiteness. For example, if someone wanted to really ram home how wicked he felt another person is (e.g. Hitler), he might say: "That monster is the Devil himself." Here "qualitativeness" is emphasized by a formal lie, and "Devil" is clearly definite.
A potential biblical example of this may be John 6:70, where DIABOLOS may be a definite noun (possibly, though I take it to be indefinite). Wallace believes that it is monadic, and, like Tyndale, favors the rendering "one of you is the Devil" (Greek Grammar Beyond the Basics), p. 265. If so, then here again we have the 'nature' of a subject stressed via a definite noun offered in the context of a formal lie.
With both examples the word "Devil" is used to stress the 'nature' of a subject, and the "qualitativeness" we infer depends on the definiteness of the nouns. If they were purely qualitative then we wouldn't necessarily even know whose "qualities" are being transferred to the subject via formal lie, as quality isn't a who.
So, both definite and indefinite bounded nouns can be used to convey the nature of a referent (="qualitativeness"), and in all cases the qualitativeness actually depends on the definiteness or indefiniteness of the terms. I would argue that this applies in biblical Greek just as it does in English. The difference is that in Greek there is no indefinite article to signal that a noun is not definite, and so this must be inferred from context.
I'm sure that Paul Dixon is a fine man with many admirable qualities, but the thesis that won him fame is not a serious piece of linguistic/grammatical research. It is an apologetic piece, inspired by fidelity to his doctrinal commitments. This is hardly surprising coming from a student of Dallas Theological Seminary!
In my next blog entry I'll list many of the nouns that Dixon claims are "qualitative," which are actually either definite or indefinite. It's quite revealing.
 The Significance of the Anarthrous Predicate Nominative in John, Th.M. thesis, Dallas Theological Seminary (May 1975).
 Qualitative Anarthrous Predicate Nouns: Mark 15:39 and John 1:1, JBL, Vol. 92 No 1 (March 1973), pp. 75-87.
 Criteria for Determining Qualitative Nouns with a Special View towards Understanding the Colwell Construction, Th.M. thesis, Dallas Theological Seminary (1996).
 A Definite Rule for the Use of the Article in the Greek New Testament, JBL, Vol. 52 (1933), pp. 12-21.
 Colwell, “A Definite Rule,” p. 20.
 See: http://www.merriam-webster.com/dictionary/expedient
 J. Gwyn Griffiths was also critical of the view that Θεὸς is used "adjectivally" (yesteryear's term for "qualitatively") at John 1:1c, and observed that, "Taken by itself, the sentence καὶ Θεὸς ἦν ὁ Λόγος could admittedly bear either of two meanings: (I) 'and the Word was (the) God' or (2) 'and the Word was (a) God.'" (The Expository Times, Vol. 62, October 1950 — September 1951), p. 315. Griffiths favored the former because of πρὸς τὸν Θεόν whereas I favor the latter for the same reason. While I'm confident that the former rendering is acceptable if interpreted properly (=functionally), I highly doubt that the Evangelist could have said or his readers heard that someone was literally Θεὸς and πρὸς τὸν Θεόν (i.e. "God" and "with God") without necessitating further explanation about how such a paradox could be true. We must bear in mind that at the time the Gospel of John was written, the Trinity doctrine did not yet exist as a conceptual grid into which such seemingly paradoxical statements could be placed to avoid cognitive dissonance. It is therefore historically unlikely that the Evangelist intended or his readers inferred what orthodox expositors tell us the controversial verse means, because if he did have such meaning in mind, then there would have been concerns expressed and disputes severe enough for us to hear about them, along with a call for discussion and clarification. That discussion, once begun, would have led quickly to the sorts of disputes that occurred only later, on the road to Nicaea.
 "The Significance...," p. 57. Dixon lists John 11:38 as the only verse in all of John's Gospel that contains an indefinite predicate nominative!