the best answers were to questions I never asked

The most useful answers I have gotten were not to the questions I actually asked. They were answers to the question I should have asked, caught by a review step I built into how I work, before I thought to ask it myself.

I used to think a good review just answered the question in front of it, cleanly and correctly. I still think that is the baseline, but the reviews that have actually changed how I work went further than the baseline, and I have started asking for that on purpose.

the gap nobody asked about

That review step exists to answer whatever specific question I put in front of it. One time, it went further on its own. I had built it to catch a certain kind of leftover branding, a name or label from an earlier version that was not supposed to be there anymore, before anything shipped. It was checking most of the relevant places. It had never checked one specific spot. Nobody had asked it to look there, and if a leftover name had been sitting in that unchecked spot, it would have shipped without anyone catching it. The review found the gap anyway and said so.

That is the shape of the thing I am describing. The literal question was answered. The review also noticed something adjacent to the question that mattered more than the question itself, and it said so instead of staying quiet because it was technically out of scope.

why I told it to keep doing that

My response at the time was direct. "Good catch. Keep being consistent with finding these kinds of gaps moving forward." I did not want the review to catch that once and then go back to only answering the narrow question the next time. I wanted it as a standing habit, not a lucky exception.

The reasoning is simple once you say it out loud. Scope limits what gets fixed in the moment. It does not limit what gets reported. A problem does not stop being a problem because it fell outside the question I happened to ask.

the same habit showed up somewhere else

A couple of weeks later, in a completely different context, I was planning a change to some of my own technology setup, and I used that same process to help me think it through. I told it directly what I wanted from that point forward. "These are the kinds of ideas I would like you to help me come up with when we are doing plans like this."

That was not a correction after a mistake. It was me asking for the same thing up front, in a different setting, because the first instance had already shown me how much more useful a plan becomes when it includes the parts I did not think to ask about.

why the narrow answer is worth less

In my own experience, defaulting to just answering the question that was actually asked feels efficient in the moment. It feels like respecting scope, but a narrow answer that is technically correct and misses the thing that actually matters is not really doing its job.

What I want now, whether it is a review or a planning session, is the literal answer plus whatever else the process actually noticed along the way. Not being asked is not a reason to stay quiet about a real problem. That is a small standard I hold it to, and it has caught more than one thing that would have otherwise shipped or been decided wrong.