I find the 'also add characters when adding words?' option very useful, it's much easier to remember a word when you have an idea of the meanings of its characters. I wish I'd had it turned on from the outset. I think this could be taken further if Skritter would also add 'sub-words' as well as characters. I suppose this is only easy to implement if the sub-words are already represented as parts of the words. For instance, you could only do it if the database knows that in 骨肉相连，相连 is meant as a single word, whereas 骨 and 肉 aren't (I think) meant as the single word meaning 'blood relation'.

Sub-words would be a handy concept to work with, but you've guessed correctly: we don't have that level of precision in our database, so we can't easily implement the feature. If we programmatically added sub-word tags, it would be hard to keep up-to-date, and there would be many inaccuracies.