The book "The Mom Test" written by Rob Fitzpatrick may help you with that. It presents ways to ask non-leading questions and understand the flow of discussion such that even your mom wouldn't be able to lie to you.

It's a short book without fluff, I recommend.

You have to decode what they say. Which is really hard because obviously working on the product yourself, you want to believe they care about your product.

For example: - "This is awesome! I would absolutely use this if it would just have this one extra feature it does not have now." --> I absolutely don't care about this product. Please leave me alone. I have better things to do.

> but I could see that none of them opened it again after their initial test.

I think OP agrees w/ you and this is their key point -- you can ignore everything they say and just look at how, or whether, they use it. More generally, I think it helps to try very hard to get at the underlying problems people have, and try to make those problems go away. People will use very terrible software (interfaces) if it solves a real problem for them. I think your signals are good generalizations to be clear, I just think we (all of us) regularly gloss over problems by focusing on tech, design, or otherwise "cool" things. It can be really hard to figure out what problems people actually have, and also whether they are significant enough to change their behavior to solve them better.