[jondeval]

I think you are describing something that is fundamentally difficult for all human beings. (1) Understanding abstract problem statements (2) solving the problem (3) communicating the solution to other people accounting for their various levels of understanding and context (4) the meta problem of knowing which problems to solve and which conversations to have.

Some people are legitimately better than others in all of the above and my view is that those who are better have a responsibility to help the team.

Here are some things that have worked well for me:

* Good information gathering can happen in group discussions but I try to avoid problem solving in groups of more than 3 people.

* A super powerful question that focuses the conversation is this: "What are the decisions that we need to make in the near term?" Now if the answer to this question involves more than 2 or 3 things, politely propose that the group focus the conversation on just 1 or 2 decisions.

* In an engineering context it is extremely rare to find someone who is both very unfocused in conversations and at the same time very effective in getting their solutions built. So, don't waste too much energy trying to correct these people or bring them back on track because they just can't do much damage. Instead spend your energy in small group discussions with other effective engineers. This is where the real decisions are made and the large group discussions are for propagating these decisions for transparency.