This is so much of an oversimplification that I don't know where to start. Let's say if you manage to simplify 'drive issues' into just two categories then there's something wrong with the diagnostics/analysis. There can be countless things wrong. If you meant to say logical vs physical issues then you're getting closer but then what appear to be logical issues (like RAW drive) may actually be just symptoms of underlying physical issues (like bad sectors). And then easy answers (to for example, get me a tool that fixes a RAW drive) are in fact your worst enemy.
Wisest DIY approach with any drive issues is given by
@labtech, don't guess, clone the darned thing. But even that's an over-simplification as you can not just blindly clone, you need to monitor the process and know when to stop even that.
Very true.
Back in the day when drives still were the common option (SSD's were new and only 32/64GB) We had to figure out fast and efficient way to get data off a drive. I used to use winfile.exe (I think it was from Win 3.11 or was it the 9x era?) Under a special PE environment (which would avoid the bad HDD being host....never run a system with bad drive to run recovery) that ignored drive security, so you could go into any folder no matter the permissions. Winfile had the ability to skip files, so I ended up putting something heavy on the "i" key (I think I used an eraser and a dead HDD to weight it down...hehe using a fallen comrade) and left it to transfer. It would do a rough first pass and skip any file it couldn't read. This resulted in a clean copy of what wasn't damaged. The rest could then be manually retried or various tools used to coax the data out, once we got the initial files out. This was my way of doing it for clients who didn't want to pay for recovery.
I used HDD Sentinel, which not only gave you SMART status but also showed SMART logs. You also would get a feeling for common errors that would indicate a drive was failing. One such failure that I never could track down the cause specifically was the drives that shows tens, hundreds of attempted re-allocated sectors but then said the area is fine. I would quickly tag those drives as bad, because clearly something was wrong, and they eventually would resurface with issues.
I also ran into issues with drives that were very, very slow. I would look up the manufacture specs and run a quick benchmark in PE, and on some drives the performance was very low, indicating an issue. Even though SMART was not triggered, once the drive was cloned and replaced, instantly the system picked up speed.
SMART is actually quite dumb, and often fails to notice issues till they are serious. Windows sometimes would give the user alerts, but in my humble opinion, because SMART often was a drive specific equation, unless the drive was triggered, it would always pass even though the drive clearly was failing. This led to many users not even aware the drive was dying and would have given them time to copy data.
Depending on the situation, a drive that had weak sectors could sometimes after a sector remap clone and have Windows repaired. As mentioned, some Windows issues like BSOD or fatal file missing errors are actually drive problems, as the sector the file was on is gone, so Windows stops. Now, if you just go into repairs thinking everything is a software issue, you're gonna have a tough uphill battle and finally admit defeat.
Drive failures are my #1 thing. I'll suspect a drive before MB or other things. (This is with a system that posts and can do a memory and basic CPU test)
@crobertson You mention diskcheck, are you referring to chkdsk? That thing is the devil's spawn, I only use it when I know I have a backup! It can destroy a drive and make your chance is recovery much, much worse!
You also mention is a drive gone to being not initialized it's bad...well, that can be true, but to flat out say that....a person could have done something and not tell you. When a drive goes RAW i'd be more inclined to say it's bad. I mean, both ways are possible, but Like Spock says: "An ancestor of mine maintained that when you eliminate the impossible, whatever remains, however improbable, must be
the truth."
For me HDD and health/data was very much an artform. You got a feel for how the system, firmware responded. Some firmware was sluggish to begin with, other systems would noticebly become sluggish with a failing drive installed. You just had to put your feelers out and see what the system was telling you, just like any other patient.
Some systems slowed to a crawl, and you couldn't figure out what was wrong. Yank the drive out and suddenly even the BIOS responds quicker. So, drive failures are not a one size fits all, but a musical instrument with different notes.