i didn't realize what i was getting myself into when i started with mame. in the last few months, i have:
(1) studied the controls of tons of games and designed a custom control panel that is as close to "universal" as i could make it.
(2) produced a good ctrlr file that makes most games reasonably playable on my panel.
(3) played many games and tweaked around their control layouts in mame, producing a bunch of .CFG files. while most games are playable with just the ctrlr file, most can be improved by hand tweaking.
(4) brought up johnny5 and made a bunch of layouts, so that i can now automatically generate an image of my control panel with controls properly labeled for each mame game.
(5) brought up mamewah and made layouts to display screen shots, my control panel images, and images of the original arcade cabinets and control panels.
(6) in order to work around various mame problems: obtained a binary distribution of mame analog plus and a source distribution of mame, applied some tweaks to mame and rebuilt it. integrated these with mamewah and johnny5.
none of this is spelled out anywhere; every step of the process requires at least a few hours of web research. most of it is an iterative process. as i've gained experience with all of these software elements and the games themselves, i've had to go back and re-do various elements.
all told, i'm sure i've got well over a man-month of system integration invested in this project (and i've only just now reached a "release 1.0" level of completion). i'm primarily interested in the end result, not the process, and i would have loved it if a vendor already had an integrated solution that plays as many games as well as my system does, at a reasonable price.
in my opinion, this kind of system integration would be a great way for a control panel manufacturer to add value.
this level of integration presupposes a standardized control panel layout, which won't satisfy every potential customer... but i think it's possible to make a very usable layout that could satisfy a large fraction of potential customers. one layout could support a fair amount of configurability, simply by allowing users to selectively include or exclude individual elements, or make substitutions between compatible controls under controlled conditions.
in my (obviously biased) opinion, my control panel layout (see the link in my signature) could serve as a great starting point for such a universal layout, and anyone is welcome to use it, if they like. or perhaps someone has a better idea. but if there were some level of standardization, i think a lot of people would make contributions of .CFG files, control panel viewer layouts, etc., which could make this hobby a lot simpler for many of us. (and maybe help you sell a lot of hardware!)
i'm not suggesting that you would bundle and distribute mame, roms, etc. rather, you would distribute config files, layouts, etc. and give people a fairly simple recipe for downloading and installing the elements they need, so they could get a very slickly integrated system up in a few hours, instead of a few months.
another possibility would be some sort of community-supported effort. for now, on an even smaller scale, if there are any individuals out there who would like to use my layout, i'd be happy to share what i have with you and receive back whatever work you put in.