Is 100 mutations in 3500 years based on the full Y chromosome, or 20% of the Y chromosome?
I think yfull sequences only 20% of the Y chromosome, so this would imply 500 mutations in 3500 years. That is 3 mutations for every 21 years. If you look at the available data it comes out at 2.5 mutations per generation, but that doesn't take deletions/insertions into account.
The best way to test your theory is to post 40 pictures of men with similar autosomes who belong to 4 distinct Y haplogroups. Then ask people to pair males who they believe to look similar. Obviously this will not result in 20 pairs, but people should be able to find at least 4 pairs.
If Y haplogroup is of no influence they will be correct 25% of the time.