This amounts to a non sequitur. Of course a change in subject would have been marked -- the syntax in 1:1c emphasizes that θεός is predicate, but only context determines if it can be definite (which I think it is).In John 1:1a ο λόγος is the subject.
In John 1:1b the subject is ο Λόγος.
In John 1:1c, why would the subject change?
If it did change, there are discourse markers for a change like this.
So even if θεός at 1:1c was articular the subject would be obvious.
This is particularly true because the three clauses are connected by και which signifies an addition to what was said before.
This is true for both linguists and grammarians.
Again, predicate nominatives are normally anarthrous, introducing new informaton about the subject. One effect is to clarify that the difference between the subject and the predicate. I would not with Harris say necessarily that having both as articular means that they are convertible, rather that the identification between the two is absolute.