I only get those numbers if I remove からかう, in which case the first page is almost filled with dictionary entries for the archaic あつらう/あとう. Also, some of the results for 誂う contain 誂える instead. Still, からかう clearly accounts for a lot more than 1% of the results. I'll revise my claim to "hardest to measure the rarity of".
Edited: 2016-01-16, 5:20 pm
On the other hand, the problem with searching for "誂う" からかう is that you only get results that contain the word written in both kanji and kana. Not to mention only getting the dictionary form. It's a lose-lose situation I guess.
This isn't unnecessary, but you start running into some really odd kanji and unusual readings (at least in comparison to modern use) when you read 漢文. Here are some examples from the beginning of Zhu Xi's 大学章句:
所以（ゆゑん） - method
蓋し（けだし） - certainly/probably
自（より） - from
莫し（なし） - does not exist (e.g. modern ない)
稟（ひん） - inborn talent
閭巷（りょこう） - cities and villages
Fortunately there are modern editions that have the 読み下し and a lot of furigana so for major works like the 大学 it's not necessary to look up and know all these kanji yourself, but it is interesting.
That reminds me: 未知生、焉知死 (voiced いまだ生を知らず、いずくんぞ死を知らん). I've seen 天稟 and 蓋し (but no more than twice), and I don't consider 所以 especially rare (as ゆえん, of course, and the meaning is different).
齎す もたらす - the font wasn't helping, I could only recognize 亠 and 貝. The enclosure is an old form of 斉 as in 書斎.
努々 - you'd expect ゆめゆめ to be spelled 夢々.
Edited: 2016-01-17, 5:29 pm
いずくんぞ can also be written as 悪, which I always think is odd.
At least with 莫し its use is widespread and is easy to differentiate from 無し and 亡し. However with いずくんぞ there are myriad of kanji with this reading and most of them have secondary meanings (安焉奚烏惡寧). I tend to wonder if there is a meaningful distinction between several of these, except for rhyming reasons (like does the basic meaning of 惡 contribute to the meaning of the sentence when read as いずくんぞ).
Edited: 2016-01-17, 7:34 pm
I doubt it; it's probably just sound. Especially in the older literary Chinese writings like the Classics, there was a lot more freedom to substitute characters that had the same sound (like the frequent use of 説 for 悦)
You're probably right. I was thinking about how the basic meanings of 教 and 令 etc. can have a lasting impact on the meaning of causative sentences. But that's a different situation altogether.
While I'm not yet proficient enough to actually sit down and read any works, I have noticed the usage of 而 and 女 in place of 汝 爾 or even 若 (when read as なんじ) in older works. The joy of trying to keep all the balls in the air when starting out.
I love the unnecessary use of rare/obscure kanji, makes things more interesting to read.
Yeah, I've been watching Kaiji lately and have been pleasantly surprised by stuff like 檄, 傀儡 and 敵愾心.