RE: Find as you type for navigation

> But it works well in more simple use cases, and it is certainly 
> the most easy to understand and use for any user.

My experience is that the other method works equally well for simple use cases -- like application menus.

Further, I do not find the arbitrary applicability in the Windows environments to be easy to understand.  It is needlessly confusing.  File lists in Windows Explorer use first-letter-only-check-where-you-are-repeat-as-necessary whereas email lists in Outlook uses type-as-many-letters-as-you-like-and-jump-there.  Do you disagree that these UI situations are similar?

Jesper, you are a big proponent of consistency and simplicity.  I find it a little surprising that you are not willing to acknowledge the cognitive load associated with determining (or guessing) from context which find-as-you-type modality is in effect.

With regard to lists of links, how does the end user know a priori if the list will be short or not?  Should the find as you type method change depending on size?  A short list will be reasonable to navigate with any method -- since it is short!  With long lists, the simple first-letter-only algorithm can be virtually useless.  It seems to me than the more sophisticated algorithm with wider usage case would be preferable.

There is a challenging issue associated with making type-as-many-letters-as-you-like-and-jump-there useable for slow typists (be they assistive technology users or not).

I am not arguing that first-letter-only is *never* better.  I am asserting that, even for short or trivial lists, reasonable people can disagree about its inherit superiority to other methods.

I would love to see a rigorous investigation and comparison of the two methods.