Great comment and we see the same tension in many other collaborative areas. When designing software interfaces we deliberately use wireframes to prevent people

Great comment and we see the same tension in many other collaborative areas. When designing software interfaces we deliberately use wireframes to prevent people from trying to add detail prematurely. But the place that bothers me the most is when writing papers. I can't find an approach that works well for all teams, or even for myself working alone - do I write the experiments and results first, or do I try to shape the overall message first, ...