Feature development is rarely so cut and dry that you can hand a developer a task and let them run with it.

To get there, you need a confluence of context and expertise from several domains:

- what problem needs to be solved (user story)

- what options are available (interaction design, technical capabilities)

- what the cost of implementing each option is, and the opportunity cost of each level of implementation / each option (technical capability, resource management, sales, user research)

- managing group consensus on the path forward (communication to technical and non-technical audiences)

- break down of any large chunks of work into smaller tasks that can be done and planning the work to be done in series or parallel (resource management, technical capabilities)

Finally, after all of that, you have a task (or several) that can be handed off.

There's really no way to get here without at least some thought into the implementation details, as the business can't make the decision on which options without knowing rough timelines.

Excellent, this all can be written, thought over and discussed in written form. I do not see how this all requires a meeting.