The problem is, people's words are not reliable. You can ask 100 people "would you pay $5/month for app X", and maybe 25 will say they would, but only 3 actually will when you build it.
There only real way to avoid building a useless MVP is to ask to pre-pay for a product that doesn't exist yet.
Just asking in passing will mean they have to digest that app very quickly and then try to make a decision. So that's not really surprising.
I found more benefit just focusing on problems, like "What annoys you at work related to your job"? Unless they just can't articulate, you quickly find things to latch onto where a technical product (software, hardware, whatever) might help and you go into more detail with them.
The goal is really finding under-served markets, which I think what the article says to do, because products seem pretty straight-forward once you know those problems exist.
Even that doesn't work if the product is in any way complex.
In someways I think it is better to concentrate on value to the purchasing decision maker. Provide the decision maker with real value below the CAC and you have a valid business idea.
There only real way to avoid building a useless MVP is to ask to pre-pay for a product that doesn't exist yet.