Designing in the browser
For developers static comps provide a confusing, incomplete and often frustrating blueprint. It’s hard to get an overview of a site from images of a subset of its pages. A static comp does not describe well the purpose and potential re-usability of its contents. There is often confusion over what things should do and where they will be reused. Developers can also become frustrated with the assumptions contained in the comp.
I say Bullshit.
The trick is communication. Designers and developers have to talk with one another to get the best results. I strongly believe that you get the best product, regarding code and design, if designers works in Photoshop and developers work in the browsers and they communicate constantly about everything that comes up.
‚Designing in the browser‘ might have some advantages for the ‚client‘, but I think the ultimate goal crafting a website shouldn’t be to please the client, but dobthe best for the actual user.
But I have to agree with O’Riordan that you often should make final detail decisions while viewing the site in the browsers. Especially things like font-size or font-weight often render differently in Photoshop than in the browsers.