Maybe it's just because of the great resources here, but code tracing hasn't really been a problem for me. Yeah, it may just be a simple "Right bank lean" or "ECT out of range" or something like that and not tell you WHY it's reading that way and actually makes you troubleshoot a bit, but 9x out of 10 for me it's been a mechanical issue (usually something not hooked up right after fiddling with something else nearby) rather than a just bad sensor. Maybe that only gives it the slightest edge of finding where a problem is over a carb, but it's still an edge.
To me, the biggest problem with the factory 80's/90's efi is a new(er) problem: the lackluster quality of modern aftermarket parts (tfi and other distributor parts especially) and the unavailbility of OE parts. Again, another industry wide problem that affects anything old enough to not have OE support regardless of what type of fueling it uses, and regardless of if said part is electrical or mechanical in nature.
And I've seen a few guys in here with aftermarket efi that praise the carb glory days, and most seem to be running megasquirt and it makes me want to ask "Why?" Sure it's cheap, you can build it 100% on your own, or just buy a plug-n-play (which, other than the datalogging capability and laptop tuning is actually a bit of step backwards since anything before ms3 can't do sequential fueling), but from what I've seen, by it's nearly open-source nature, the support and feature implementation is just not there like it is for systems like Holley, BS3, Fast and even the basic TBI systems like Fitech (whole 'nother can-o-worms there). And I know the MS stuff is probably more popular for the kind of crowd we have here than any of those, but I also see way more unfinished MS systems that end up going to a back to a carb or a modded factory efi (or just plain giving up and selling the whole project) than I see of if happening with the "big names"
