I recently got around to reading Breaking into Japanese Literature and I've noticed a few things regarding irregular okurigana and kanji readings. For example:

坐る vs. 座る
度 vs. 毎 vs. たび
確に vs. 確かに
潤 vs. 潤い
真黒 vs. 真っ黒
色沢 vs. 艶
etc.

These are all examples from Natsume Soseki's short, The First Night. I was just wondering if the above irregularities are due to the fact that the writing is dated (late 1800s, early 1900s) and thus before some of the major post WWII reforms in writing conventions. Also, how often do contemporary authors deviate from the standard for purposes of style and effect?

It varies; I think most authors don't have particularly non-standard kanji usage, but some do. For instance Kyougoku Natsuhiko's "Ubume no Natsu" makes use of pre-simplified kanji forms and uses kanji for 'grammar words' like せい and はず where most texts would use kana. I've also seen some old-style kanji usage in a detective novel set in the Edo period; my guess is that in that case it's intended to add to the period flavour.

The only places I've seen real pre-war kana usage in the "write 思う as 思ふ" sense are some aozora bunko texts and some of the chapters in Miller's _A Japanese Reader_; the latter was printed in 1962, at which point 旧仮名遣い is described as "old-fashioned" but still worth being able to read. Fifty years later I'd say it's now dead unless you needed to read period texts. My modern copy of Tanizaki's _In'ei Raisan_ (written in the early 30s) has modern kana spelling but seems to have retained the original kanji choices and okurigana, as well as the angle-bracket style "multiple kana repeat mark" punctuation.

Okurigana variation is probably more likely on the Internet due to randomness -- I don't think that has anything to do with author style, really.