i wonder what the charity's overhead is, compared to the theoretical "loss" rate to addiction/booze/whatever we want to decree an ineffective use of someone else's money.
Looks like administrative costs are about 10% of overhead (fundraising is a net positive, so I wouldn't hold it against them).
Disagree. Fundraising is net positive for the financials of an individual organization but not for society. Donors have a finite amount of money to donate to charity so convincing them to give it to you often just means they don't give it to someone else.
And even in the alternative that the donors would have spent the money in the general economy, that has obvious societal benefits as well, e.g. job creation or easier access to capital. It is better to have half as much money going to charities that spend 90% of it on programs than to have twice as much going to charities that only spent half of it on programs, because in the first alternative the other half of the money is at least going to something potentially productive rather than a zero sum competition between charities for donors.
That's not always true. Sometimes a potential donor doesn't give at all, so if you convince them to give, then this is a donation that would not otherwise exist.
You're making an assertion that needs some citations to back it up - on a prima-facie view, wealth not spent on charities generally sits in coffers or is used to make more money, which does nothing or little directly for the poor.
The view isn't whether the money is "being put to use" but of what qualitative use it's put to. For every angel investor or VC there is out there funding startups, there are 10x or more wealthy individuals who don't invest their wealth with any goal other than increasing what they have.