Currently my stats for characters learnt show:

1202 writing
585 definition
567 reading
1215 tone

Why are the reading and definition so much different?

Are you studying both Traditional and Simplified?
Skritter counts Traditional and Simplified separately for the writing progress, but not for the reading progress.

No I'm only studying simplified characters, I'm wondering if because I marked quite a few readings and definitions as easy in the beginning it just hasn't tested them.

If you're using the iOS app, or are studying Japanese, then character readings are not studied/learned during the course of studying their containing words' readings. And on both iOS and web, this is true for character definitions.

It's just writing, tone, and web-version pinyin prompts that score reviews for the sub-characters in a word while reviewing that word.

What is the reasoning behind not counting word reading prompts as character reading reviews?

Because you're not providing an answer for each character, and in particular, if you get it wrong, we wouldn't know which character you missed. So if we submitted a review for each character only when you got all of them right, then you'd have 100% retention on each sub-character item that you don't see on its own.

On the web version for Chinese, where you do type in the pinyin for each character in the course of typing it in for the word, then we do submit the sub-character reading prompts. Just not on iOS or for Japanese, where there is no typing.

Ah, good point that I hadn't realized about readings in iOS. (Although mine learned counts are all even... probably because I also add characters in words).

One option for iOS would be to put grading markers under each syllable (similar to the whole word marker at the top of the screen for other prompts.)
Probably would be a little confusing to some... I didn't know the top of the screen (word) grading could be changed until just last week!