I have some PC components that have failed for mysterious reasons. I am aware of some of these failures being treated with an oven bake to reflow solder joints. I want to examine and test individual components on the PCB (resistors, inductors, capacitors) instead. What device would accomplish said task?
I've done some digging and looks like DE-5004 is a good start. Not as good of an LCR as DE-5000 but I believe, for my application, that doesn't matter. I've done a lot of soldering for RC and I've replaced numerous blown electrolytic capacitors. But I've never taken an iron to microscopic resistors and diodes.
Any pointers, comments?

The most effective way to debug electronics circuits is with a oscilloscope and follow signal paths from part to part and see where the signal disappears.
That part is likely the problem.

It is prety rare for ANY meter to register digital or ananlog signals in a circuit.
How do you know what the meter should be reading?

Do not waste your time looking at a signal unless you KNOW what it should look like at that point.

It's really hard to do without good knowledge of electrical engineering and a diagram of the board at a minimum.

You don't even know what the capacitance of chip caps are supposed to be, or inductance, or identifying components, etc., as most stuff isn't labeled, or it's labeled in code or custom part numbers.

Usually if you have to ask "what device", you don't even know what you don't know to even start.

Generally with a knowledge of the basics, caps and obviously blown components are all you can look for.

A well done board will have enough marking for component positions to allow you to work from a schematic of the circuit.

Many modern boards want to save space and so there may be no component markings actually on the board.
You then must work from an assembly drawing for the board.

For commodity commercial equipment even schematics can be hard to come by.
And if there are ANY field programmed parts on the board you need their schematics or programs.

It can take hundreds of pages of VHDL to specify the design of a large Field Programmable Gate Array.
Signal processing may have taken a 'C' program, turned it into VHDL, and then programmed the part to implement the 'C' program.

It takes a huge amount of work to debug these designs since the error could be at any stage of the design and even in the board layout itslef.
Finding errors in something operating at GHz speeds is a special nightmare.

Most boards I repair on don't even have a silkscreen. Depends on how secretive the oem was being.

Single/double layer boards are easy enough, but once you have more than that, it's a nightmare.

We used to routinely use 10 and 16 layer boards for compact high speed design.
Putting a soldering iron on the board without preheating the thing to at least 400 F or so could cause cracked plated through holes.

The most expensive board I ever made was a 16 layer teflone board for high speed digital and RF.
It ended up at around $23k a copy.
It was 10 inches wide and over 24 inches long.

It provided the interconnection between a VME chassis full of RF boards and the DSP chassis that sat behind it.
Each U-2 system carried four of those boards.
How many RF channels do you want?

That is well and beyond what most DIY or even light pro guys would even see in their lifetimes.

If a guy is asking what tool to inspect resistors, caps, etc., he is just not even in the same ballpark.

I usually work on 4-6 layer boards, but like I said, no silkscreens on 90% of them. MAYBE labeled with individual R1, R2, etc. If you don't know the circuit schematic (which consumer goods don't have available), and including any custom PNs on ICs, 99% of folks are beyond their ability.

Actually, my main job is reverse engineering boards like I described: no silkscreens, no schematics, custom part number ICs, obsolete components, etc. It can take a LONG time. Just here is a board, figure out how it works type stuff.

ETA: That is an expensive fucking board...

We used to 'update' bards for older Naval Equipment.

Make a form-fit-functional replacement with newer parts that are available.

The -55C to +125C and hermetic packaging is not a very large market anymore.

Think smaller cards (4 in x 5 in) often only double sided but with 24 16 pin DIP parts per card.
In hundreds of varieties and functions.

They must still at least give a 'PASS' in antique automatic testers also.

Most ended up with a single FPGA on the board.
A few would need additional signal conditioning (level shifting, additional drive).

And almost always a lot more anti-static protection.
A lot of CD4000 family stuff out there.