I think the point being made is that giving your client the option to determine what your work is worth carries with it the hidden threat that you'll lost inter

I think the point being made is that giving your client the option to determine what your work is worth carries with it the hidden threat that you'll lost interest in doing work for them if they undervalue you. Only useful with existing relationships, but it's like getting validation that they value the work you've done. The real problem I see with this is that occasionally clients will opt to do something fairly complicated from a technical perspective that will end up having a very small financial return. If the end results only improve their situation by $1000, but implementing it cost you $2000 based on your usual billing rate... well you're not going to get your money out of it. This means you need to be anticipate it and be picky about what projects qualify for this type of pricing. You need to be able to validate ahead of time that the return will actually be worth the amount of work... which is a really good idea anyways, because it means you have to wrap your head around the actual value of your end product and really understand the problem you're solving. It increases your involvement with the client's business, which many clients and even some developers don't realize is always necessary in order to deliver a quality product.