The fact that your article starts with a fallacy (the bayonet training part) doesn't help its credibility...
It is also slightly off topic. The original article talks about catching criminals, yours talk about conviction. The issue is about what constitutes rape in Japan, not the effectiveness of the Police.
There is still an interesting bit in the original article : "The hunt for things to do may sometimes be beneficial. The number of reported cases of children being abused at home has almost doubled since 2010, despite the declining birth rate. That suggests the police are increasingly intervening in the domestic sphere, which they used to avoid.". It means that we might see an improvement when it comes to domestic violence (which can include rape).
I would caution anyone reading this top-voted comment to look at any other story about crime stats in Japan. This knee-jerk response about manipulated statistics is almost always in the response. If you don’t think crime is exceptionally low in Japan, go spend a year or two outside Tokyo and get back to us.
I lived in Osaka, which is ostensibly the crime capital of Japan, for a few months. I have never felt more safe in my entire life. Yes, there were the characteristic Yakuza businesses, but I would have been surprised if I dropped my passport at 2 am and it wasn't returned.
I grew up next to (literally) the Yamaguchi-gumi headquarters. Never once did I feel unsafe there, or anywhere else in Japan. It really is a night and day difference.
