I'd say in the past year I've allowed about 12% of my holdings to drop. Given my total holdings that's not an inconsequential number.
Looking forward I'm considering ever a larger clean-up operation.
Concerning my "personal drops" :), I'm sometimes amused, rarely irritated and typically some degree of happy when someone pick up one of the domains I allowed to expire. Happy because I knew it was a decent domain, just not a domain up to my current standards. When I see a prompt re-reg I'm glad someone else secured a decent property. Also, I get a tiny bit of validation out of the re-registration, like "See, I knew I was on to something".
At least the drop and re-reg confirms that I'm not the only deluded domainer. :P
The tricky part is the selection criteria. What to drop? Why?
I look at traffic. I look at income earned via parking. I look at enduser value of the traffic, since PPC payments are often out of sync with enduser "converted sales lead" value. So I may hold onto some domains that didn't pay for themselves with parking revenue because I perceive an inherent value in the domain and know there is some demand.
Example: A legal domain that may only receive 5-6 type-ins a year, 2-3 clickthroughs. I know each clickthrough could be worth $$$$ to a lawyer so - I wait, patiently - until my lawyer buddies wake up to the reality of competitive marketing in the Web era.
So, what percentage have you allowed to drop during the past year?
What are your thoughts going forward? Plan to refocus? Plan to cull even more?
What criteria are you applying?