As Rachel said, the session on How Chemists Search was great. We were seated right in front of some actual (retired) scientists--in persona speak, they're the "disengaged advocats"-- a misnomer as chemists are very engaging...and opinionated!
Aline and I took the opportuity to pick their brains for thoughts on searching. They said they like google. If they had to design a search interface for ACS they would do what google does--they'd have links above the search box so they could specify whether they wanted to search journals, C&EN or society content.
Interestingly, these guys were very open to using collaboration tools. They said that when they were starting out, it was easy to know everyone in your field. Now, they think it's much harder to do that, so online collaboration would be something they'd use. Most interesting is that the data element they want the most when making an online contact is an email address. This mirrors what we learned in San Francisco about the Member Directory...people want to find people online, but they immediately want to establish a personal contact.
In another session on Monday, we learned about the effects of uncertainty on search. A key takeaway from that session is that the search results interface must enable easy and effective searching under uncertainty. The search results should suggest possibilities and allow people to follow tangents.
One final comment: Librarians are passionate about their work!