We do a lot of computer troubleshooting and diagnostics for clients. Macs (not MACs BTW) probably account for around 1% of our workload. I think there are a few reasons for this: they are a slightly smaller market to begin with; users get sucked in to Apple's product life cycle, so systems tend to get replaced before they're old enough to have common failures; Apple (maybe) has better manufacturing controls, so there is a lower percentage of early failures; and, probably most importantly, Apple has its own stores everywhere, with a strong brand, so Apple owners will tend to just take a misbehaving system in to the nearest Apple store.
In our case, we have separate hardware for testing drives and memory. That pretty much leaves the logic board and/or video board, which are integrated on a lot of Macs, and experience gives us enough hints on the rest.
So, I don't mean to discourage anybody that wants to build a really cool tool for MacOS diagnostics -- I'd love to see some of the old-school hacker mentality come back to the Apple market (e.g. MacsBug) -- but there probably isn't a lot of demand for it.