That's an interesting question to me, just because I've seen so many frames fail there at the drive-side chainstay junction.
The socketed dropouts I've seen, as on my Trek 720, do seem to maintain the full width of the dropout up to the socket plate, and it's all of one piece, vs. the stepped-down toungue of many dropouts where the cross section is supplemented by the tubing, right where they so often fail.
I suspect that using socketed dropouts means extra hassle and inventory for the builder, especially for the smaller frame shop. Getting the diameters of socket and tube closely fitted means adjusting the taper diameter and cutoff point to the socket size and will probably always mean more parts-selection scrutiny, calculation and ultimately, more inventory.
I've seen many very good frames use socketed dropouts, but can't say I find it to be aesthetically pleasing. I bought into Trek's advertising back in 1983 or so and so have always looked positively on that marginally on-topic fabrication. It has proven flexy yet so far durable.
David Snyder Auburn, CA
----- Original Message ----- From: "Don Wilson"
> I've always wondered why drop outs were attached
> tongue in groove rather than by a lug (i.e., a stay
> inserted into a cylindrical end on a dropout). A lug
> would seem a stronger joint and, could likely be
> engineered and machined to be quite light. Fabrication
> seems technologically feasible. Assembly seems no more
> difficult, perhaps even easier. A lugged bike might
> look more unified (and beautiful) in appearance with
> matching lugs at the frame ends and frame joints,
> rather than tongue in groove at the frame ends and
> lugs at the frame joints? Perhaps the KOF frame
> builders could weigh in here and explain why tongue in
> groove is preferred. I assume it must be superior for
> some reason, or it would not be done, but I'd like to
> understand why?
>
> Don Wilson