Despite the fact that what you say is pretty rubbish: unless someone played with the functions that are called to look at files during a scan, it's pretty unlikely to have any effect; I get the knee-jerk reaction to having the latest copy to attempt to weed out people who are sub-par at debugging.
What exactly am I saying that you consider rubbish?
Also, I never said the latest patch would fix your problem. But one never knows unless one tries. It's always best to have tried the latest version, updates, and patches one reporting issues.
If I'm dealing with software that is (say) a year old... or a major version back and I'm dealing with a complex system with 100's or even 1000's of changes over that time, then it's very valid to say "checkout the new version" ... because the number of changes is just so vast that nobody truely comprehends it anyways and the people who are closest to comprehending it are a year away from the code.
MB's development cycle is quite different. Bugfixes for any number of things are introduced in every patch, usually daily. I'd say in three weeks, there are probably at least a hundred code changes, covering a massive variety of bugs and features. That's just how Steven works, which is why we always recommend to upgrade before further troubleshooting.
Regardless of the differences in Steven's development and release style, debugging can only easily be done by him. There's no source tree or code available publicly, at Steven's choice. If you want a debug version, change the topic title to something more descriptive of your problem, then post again in it asking for a debug build.
Bold words in my posts are links unless expressly stated otherwise.