If you are using MHR for parity, then what's the flaw? In your examples MHR worked. You beat the teams your were favored to beat and lost to the team you were expected to lose to. So the fact that one win might of moved your MHR rating by a hundredth of point matters? The same win hurts you in Krach.
A win against a bad team still hurts you in Krach, Krach just hides it. In the CHN power rankings (which use Krach), both Boston College and Boston University have played 17 games. BC has a better record at 13-3-1, but KRACH rates BU higher at 12-4-1. The difference is the teams that they play. Wisconsin is even worse, at 18-4-0 and rated lower than BU and BC. Your krach is based of the krach ratings of the teams that you play, so playing a team with a lower krach will ding you even if you win (Just as it does with Boston College). The only difference is that Krach doesn't have an easy to read explanation of exactly how much the win against the crap team cost you....which is why no one is complaining about it on here.
In MHR, you can precisely determine ahead of time how a win against a bad team will affect your rating. If you care that much about your rating, you could change your tournament plans. In Krach, you know it will impact your rating negatively but have no idea beforehand exactly how bad it will affect you. Although, it's all kind of a moot point for you since HPR doesn't rank A teams...