As someone who has self-published books that no agent or publisher wanted to touch, I still maintain that self-publishing doesn't count. I mean this in the sense of self-publishing proving nothing about the readability or quality of the book. It's merely a statement that the writer believes in their own work. Some reasons:
1. The traditional publishing industry, for all of its faults, at least publishes book that are recognised as being of a minimum publishable standard. This means the craft is proficient, the MS mostly error-free and – most importantly – that the book has been 'certified' as having a structure that works. It is a professional product. Self-published books need not meet any of these standards, though sometimes they do. And while it's true that a lot of dross is published conventionally, I'd say that 90% or self-published work is sub-standard. There's a difference between being a bad writer and not being a writer at all.
2. I've seen work by a lot of self-published writers (some of them successful) and I'd have to say I've not yet seen one that can write according to the most basic definition (an understanding of long-form narrative structure, scene building, characterisation and even basic grammar). They've uploaded books, but they can't yet write. I think that's bad news for the language in general and worse news for the writers themselves. Evidently they've dreamed of being published for years and have worked hard to produce a book, but they still haven't reached the minimum publishable standard. Isn't that what they want?
3. I suppose what I'm saying is that – initially at least – a third party has to judge that the work is proficient. The writer himself is often the worst person to judge this, especially at the beginning. I self-published my first novel after I'd had three novels traditionally published, none of them seriously copyedited by the major publisher who accepted them. My work has also been praised by a Booker Prize-winning author. Such things gave me confidence to think that I might know what I was doing. [They didn't want the fourth book because my overall sales meant I was a dead weight on their bottom line.]
4. Why did I self-publish at all? Because I'd written the books anyway and nobody was going to publish them. It seemed a shame to leave them on the computer never read. As it happens, almost nobody has read them anyway because I've never done any publicity for them. Why were they rejected? Well, all of them were requested in full by a number of agents or publishers, but all were turned down because of their subject matter, genre or tone. I do have a talent for writing books nobody wants to publish.
5. I've had more books traditionally published than I've uploaded. And I have another two novels on the computer that I couldn't be bothered to format for self-publication. I've realised now that I am writing for myself only – to improve as a writer. I will continue to send each MS off to a publisher or agent, and I will continue to write if they are rejected, but I won't be self-publishing any more books. I don't want to be part of a market in which nine out of ten authors are amateurs (by which I mean their work is not of a professional standard). I don't tell anyone about my self-published books because, for me, they don't count.
I suppose it depends what you want. I want to produce writing of a professional standard more urgently than I want to be read. I can understand that other writers want to say they have 'published' books more than they want to say that they are good writers.