Are documents of a certain sort needed for takeoff meetings (Two Takeoff Meetings)?
I'll speak of the "Manifesto for Agile Software Development" (2001), both the values four values 
and the less influential twelve principles , as well as the written conclusions of LAWST1, which appear as an appendix to Cem Kaner's "Improving the Maintainability of Automated Test Suites " (1997).
Both documents emphasize the **in-the-trenches credibility** of the authors. Manifesto: "We are uncovering better ways of developing software by doing it and helping others do it." The LAWST document was from "thirteen experienced software testers."
Both are **punchy**, presented as lists. The LAWST document is considerably longer than the Manifesto, but the memorable bits are at the front. In the Manifesto, the memorable bits are the four values on the first page.
But both have more non-memorable text that memorable text. Does that have a function? I suspect they provide what I'll call **concrete credibility**. The "values" page of the Manifesto is not obviously actionable, but the principles page has more of a "do this" air. (Origin of the Manifesto Principles) The main takeaway of the LAWST session was "don't use capture/reply tools" and "black-box test automation is way harder than you probably think," but that's softened by discussion of things you ought to do if you embark on test automation.
The documents strike me as simultaneously **negative yet forgiving**. They're trying to overthrow an orthodoxy but want to give its purveyors credit for good intentions. (In both cases, I think the authors were less forgiving behind the scenes.)
* The Manifesto is clearly outlining a radically different style of software development but goes out of its way to say "while there is value in the [conventional] items on the right, we value the [novel] items on the left more."
* The LAWST document doesn't come out and say "black box test automation is probably a waste of money outside of uncommon situations," though I suspect that was the sentiment in the room. Instead, it offers a list of considerations for people going down that path.
I have a vague feeling that both documents have a memorable core surrounded by text that functions more to increase believability than to impart information (whatever the original intent).
(The core of the Manifesto is obvious even to outsiders – it's the first page. The core of the LAWST results is more obscure, captured in phrases that only stand out to Craft insiders: "Common automation initiatives failures are due to using capture/playback as the principle means of creating test cases" and "Straight replay of test cases yields a low percentage of defects" and "most bugs are found during manual testing." These were *not* the common wisdom at the time.)