In graduate school, I spent (wasted!) much time trying to prepare and send out articles to academic journals. I'm not saying I didn't learn anything doing this, I'm sure I did, and I did publish a couple of articles by the way, but it was not till the end of graduate school or perhaps even the beginning of my professional teaching career that I really understood the difference between a publishable article and a highly competent, even potentially interesting and original article-length piece that is nonetheless not an article and will be very unlikely to be accepted as such.
It has been very much the same thing for other kinds of book, only with a much longer and shallower learning curve. At age 38 I now feel I suddenly understand what makes a really viable commercial book proposal or book project, something I truly haven't felt in the bones until quite recently. I think that while reading and writing are a very good (the only!) way to learn craft (sentences, paragraphs, storytelling), the editors and agents are likely to learn much earlier in their careers than the writers what makes for a good pitch, in the serious rather than the trivial sense.
(But then their job is to recognize them, not so much to produce them themselves; it's just a different skill set, that's the long and the short of it.)
I have always been the kind of person who assumes that I will never know more in future than I know at the present time. (I am not sure if this is pessimism or over-confidence.) It is nice to know that I am capable of learning new things...
[ED. AFTERTHOUGHT. This sent me off to see if I could find a version of my first real published article online, and indeed it is here, at least the abstract and opening paragraphs. But in fact this piece was based on the senior thesis I wrote as an undergraduate - this one! - and the other two articles I published in graduate school were taken from dissertation chapters, not from seminar papers. What I learned from this: it is relatively unlikely that a paper one writes for an actual class is going to turn into something publishable, because its argument is almost always constructed within the professor's framework, and would necessitate wholesale re-framing in order to be suitable for publication, even if such a thing were possible, which it is usually not. Publishable work is much more likely to come out of the context of an independent project - this is why I am often suggesting to graduate students in the humanities that while publication is quite important, it makes more sense to wait until you've drafted some chapters and then see what couple pieces should come out of those and go out to journals rather than spending a lot of time polishing seminar papers when it really would be more valuable to make tracks on the original work of the dissertation.]