What if it is for a good cause.
What if it is to help the children.
What if it is to help our Vets?
When is Fraud Waste and Abuse OK?
When does it lead to a manmade disaster?
From an absolutist position, WF&A is never ok. From a realist's position, WF&A is inevitable, you can only work to minimize it.
There are always questionable decisions, as the child food bank example illustrates. If there were a million kids being fed, then I think you would be able to say that WF&A has been minimized. If it served 2 kids, then the WF&A is astronomical.
In other words, part of the equations is efficiency of the program. Regardless of the policy, you want the most efficient program possible. That reduces / eliminates waste. Hiring and or using honest employees and volunteers should minimize fraud and abuse. But you can never get to that perfect level of zero WF&A. For example, going back to the food bank example, an employee takes an extra 3 minutes reading the paper while on the toilet. That is a form of waste, but the vast majority of people can live with that.
The real question is how much WF&A we can live with for a specific program. If you take Medicaid, fraud rates are estimated to be over 10%, but we are not doing anything to fix the problem and are not looking to eliminate the program. If we can live with $70 billion in WF&A in a single program, then is there any real ceiling to what we will tolerate?
Similarly, the public seems to get enraged when $50,000 is wasted. So the answer appears to be that if the public can generally understand the number of WF&A, they get mad and demand action. When the number is too big, then they can't.
Practically, I think that the answer to your question is that you have to know what the total cost of a program is (including WF&A), what benefits you get from the program is, what trade-offs you make, and then you decide if the cost to benefit ratio is acceptable.