Very nice typing speeds. As DreymaR has said, I think you should post your speeds in the General section as well to give everyone a better grasp on its importance.

As for kana input, I can already see that it's not the best way to type in Japanese, but I still prefer kana input over romaji input. Plus, fortunately or unfortunately, I don't type in Japanese that often for it to matter.

Thanks for the suggestion. However, for the sake of consistency in tracking relative improvement, I will continue to quote numbers from http://labs.jphantom.com/wpm/ . TypingMaster is unrealistically unforgiving, although I don't mind using it for comparisons.

I must also admit that the test I use is not very accurate - it displays both cpm and wpm in the results, with 1wpm varying in value between 4.3 and 4.8 cpm! All up, it probably over estimates speeds by about 15% if using the standard 5 characters per word.

At any rate, on day 18, I am managing between 300-400 cpm (60-80wpm), with peaks of 85 wpm on the aformentioned test.

true...but the problem with the peak typing speed is that when you actually type, you do make mistakes which you have to fix and you don't type at your peak speed but at your average speed during your typing span. Thus, the most accurate speed is the average.

Would you rather be among the best devils in Hell, or among the lesser angels in Heaven?  ;)

To me, peak speeds are pretty much useless. I can of course devise a test that gives me a "typing speed" to match Ms. Blackburn more or less, but what's the point? I'd only be lulling myself into false confidence and produce results that don't compare to anything interesting really. The thing I like with the TypingMaster test is that it seems to report something related to how fast I can actually type a long text. And that's what I need to do in real life, not excel at the arstneioarstneioarstneio race track.