In the examples (Carousel & Paper) the design isn't coming from the very top. Instead, I suspect the designers were given requirements by non-designer PMs and told to make it look nice. It works about as well as programmers being given requirements by non-programmer PMs and told to make it maintainable.
I still don't know how much of that dysfunction is because creative people can't do their best work without freedom, and how much is because the briefs/specs are bad because they were made without deep understanding.
The ideal consumer software company has a great designer, a great programmer, and a great instigator at the top.
In Google's visual redesigns - and I've worked on 3 of them now - the design is coming from the top, or rather Larry (or Marissa before he took over) gives carte blanche to a designer and says "Make search look nice". There are restrictions in terms of what is technologically possible or feasible with the engineering resources at hand (steering a massive product like Search to look different or do something different is not easy), but it's not like PMs hand the requirements over the wall.
I think that the problem is communication itself. Communicating requires creating a shared vision in both the speaker and the listener; when the thing to be communicated is basically emotional (as design is), then reducing this to words necessarily loses information. And it's a low-pass filter: it loses precisely those elements that were daring, unique, and innovative in the original design, because those are the elements that the listener/implementer is least familiar with. Designers try to work around this by using pictures - or even better yet, code - but the problem is that your product is ultimately designed to be an experience, and you can't convey that experience without creating it.
Maybe the critical element isn't who's on the founding team (although having those skillsets covered certainly helps), it's that you can get all of them in a room together and have them each responsible for all of the success of the product. I suspect that a great instigator who goes on eLance and 99designs to contract out the design and programming doesn't do much better than the big company does.
It's hard to imagine Drew not caring about design, or not understanding the things you've mentioned. Dropbox succeeded over other competitors mostly because of its fantastic design and user experience. Is Dropbox simply so large now that it's becoming just another typical big company with dysfunction typical of big companies? If so, that's unfortunate. I wonder if YC could amass some wisdom about how to stave off such problems after companies grow really large.
Thanks for chiming in with an inside view, by the way.
I still don't know how much of that dysfunction is because creative people can't do their best work without freedom, and how much is because the briefs/specs are bad because they were made without deep understanding.
The ideal consumer software company has a great designer, a great programmer, and a great instigator at the top.