(This is another reason there shouldn’t be “panelist events” where half the people voting see the course on the same day - they are likely to form an inaccurate view, whether too negative or too positive.)
Tom,
Going one step further, I think panellists shouldn't be able to contact the course (or anyone associated with the course), or, tell anyone they are on the panel, and go alla Michelin (caveat, I have little understanding of how Michelin actually works, but I'm pretty sure they don't contact the restaurants in advance for a rating).
It helps your argument in a few ways: one, it removes the decision by the course on whether it's ready to be 'rated', and will give a better reflection of the overall day-to-day, and as you note above, the scores will average out in line with conditions (in the case of wind as an example).
Also, it can allow the rater to focus just on that specific day/visit. In the restaurant example, if someone gets overcooked pasta one day, it might just be a bad day for the kitchen, and not reflective of their usual standards. But one slightly lower score won't kill a course, and if it's really a T100 course, it won't matter in the long run (even if it means it has to wait another 2-year cycle). Cream always rises to the top.
It would reduce the ability for raters to see a course early, or too early, and either force them to book with the public, or find a member. It would also weed out those just looking for access, from those who genuinely have a passion for seeing new places and understanding the unique qualities that make each course special.