> That's interesting; shouldn't cost-of-input, income, expenditure measures all equal the same amount?

In theory, yes, they're all the same.

In practice, measuring each of the three has its own challenges, which means the raw numbers will differ.

It's just like physics - you can use the work-energy theorem or Newton's laws and arrive at the same result in theory, but depending on the experiment you're running, one may be easier to measure accurately than the other, and you'll end up with slightly different numbers on paper.