Think from the variety of opinions, you can tell there is no ONE answer on the right flow of the recommendation, esp with regards to the recap upfront .
In my view, it does make sense to have it, but keep it concise to 1-2 sentences max. Reason is that after 30-40min, a lot of discussion and pivoting will have happened and it sometimes gets easy to loose sight of the bigger picture. Hence, it's good to bring back the focus on what you were solving for, i.e. the case objective(s). Life is not only about efficiency.
Based on above, suggest structure as follows (quite standard, but very effective)
- Recap on prompt + objective (1-2 sentences)
- Conclusion / 1-line recommendation
- Supporting rationale (2-3 bullets)
- Next steps / challenges / key risks (of a reco)
On step 4 - here you should always pick only one of the 3 options - select, based on common sense. Some examples..
- When you do a valuation case, often no point in talking about key risks, but rather about next steps you would take to make the estimate more robust
- For market sizing, same thing.
- For a profitability case with a recommendation on how to turn around / improve, key risks or implementation challenges of recommended initiatives are often the way to go
Hope this further clarified. Feel free to reach out if you'd like to learn more on how to perfect your recommendations and/or seek detailed feedback.