I work for a VC, and it's sort-of like this. Not so much because we think so highly of our ability to detect shady people, as because the checks we do are fine
most of the time, and constantly being refined. But they do need to be cheap - it's not easy to place money and manage that portfolio for 5-10 years cheap enough to actually make a good return when factoring in investments that fail for reasons that are far harder to identify ahead of time than outright fraud.
The outright frauds are rare enough that checking for them needs to be very cheap for it not to be more cost-effective to spend that money improving other deal-flow filtering. That's not to say that VCs aren't trying to improve on it. It's just that fraud is just one of many risks, and managing risk is about taking a dispassionate view of which measures improve your returns the most.