As it became clearer by your comment to the answer of @JordonS: I think the difference is made through a notion of agency, while this difference is hard to sustain within Dennett's conceptual framework.
First, let me apply the definition to digital identity in common-sense terms:
(2) and (3) seem to be true, by mere being a representation of a person (or the conceiving of the digital identity as a representation of a person). Even if a digital identity would be generated without having a person that is represented by it, (2) and (3) would hold even if it was created and maintained by a computer.
(1) is arguably true as digital identity is created and maintained by rational entities (that could be computers) and so it is itself in some sense rational. It "acts" with regularities within judgements that can be expressed by rules.
(4) Even a computer could be programmed to parse comments etc. so that it is fulfilled independent of its creator and the problem of representation.
But what about (5) and (6)? It would indeed sound strange to say that a digital identity would itself be a language user. It is constituted by language, the creator speaks through it by language and so on. But the digital identity is not itself a user of language.
And especially (6) is hard to think of how this could work. Because a conciousness/self-conciousness would implicate a certain independence. But representation implicates a certain form of dependance on the represented that is hard to think as (self-)conciousness.
First conclusion: Using "common sense", applying will most probably result into a rejection of the idea that digital identities are persons.
Talking about Dennett:
But in terms of Dennett, it could not be freedom that constitutes a difference, because he rejects such a thing. So for him, it will be very hard to find anything independent. Everything is determite. Applied to the concept of persons, agency will loose its special feature. And therefore you have to differentiate:
Do you speak of an application of Dennett's definition of digital identity in common sense terms or in Dennett's terms?
Perhaps this is the problem behind Dennett's reductionalism: In his terms, the figure of a book is as good a "person" as we are. It would be totally determined (by the author), but "acts" like a self-concious being, moral actor and so on. So we cannot possibly know if we and our lifes are not "scripted" by a person of an "ontologically higher being" either, as we are determinate. It is philosophically poor to just say that there are metaphysical entities like this as we have no possibility whatsoever to state their truth, but there is no reason to reject the possibility of the existance of these entities either. That's the paradox that evolves out of his rejection of spontanity/freedom as part of (moral) agency.
So for Dennett, digital identities could be described as persons, following his definition. For the common sense, I think it is counter-intuitional. So you have to develop a positon of your own on this. Do you follow Dennett in determinism? If so, will there be any chance of not applying social personality in a full sense to digital identities?