100 to 200 microns is the same as 0.004" to 0.008".
3D printers using plastic thread do what they do well. Every machine, every process has limitations and hazards, and with a 3d printer, the hazards (depending on what setup) include fumes, UV and low-power (but still dangerous) laser emissions.
The current major weakness of inexpensive 3 printers- the choice of materials... because of the thermal nature of how it has to fuse, the end product isn't going to tolerate high temperatures well, nor will it have a polycarbonate-level of strength... or high UV resistance. Basically, your part is limited to the material's limitations. This WILL obviously change a bit with time, but the fusion process really is the determining factor of material.
That being said, it's not the end product of the 3D print that is the end-of-the-story. Sure, we can print out a replacement part for some plastic thing that cracked and failed... and what we 3d print might not last a quarter of the time of that original part... but:
The 3d printed aspect means that we can make PATTERNS for other processes... like molds for pouring metals. The finish quality isn't as nice as a hand-made wood pattern set built by a veteran craftsman of the 30's through 60's, but plastic is easy to clean up by hand.
I got a buddy who's been experimenting with 3d printing parts, then burying them in sand, and using it like 'lost foam' casting. He hasn't gotten it quite right yet, but he's still tryin'. Most successful so far was a 3d model of a large timing gear... about 6" diameter, 2" thick... and burying it in greensand... with a sprue and vent molded in. When you pour aluminum in, the plastic melts, but it doesn't 'come out' of the mold fast enough to accept metal, it cools the metal too fast, and leaves big gas bubble voids where it burns. So his tactic changed a little... he's been trying to use weaker material, and extremely thin (think totally hollow) printed patterns, then BAKING the mold at just high enough temperature to melt the plastic out as he's pouring aluminum.
One of these days, there'll be one in MY shop, but what I'm seriously thinking of doing (once my new building is up, and the big CNC gantry is operational) is fit a wire feeder to the Z2 axis, and go with really SMALL wire, high-speed feed, and set the voltage/current level so that a MIG nozzle becomes the same as the plastic squirter, and see if I can 3D print a part using ER70S. If it works, I might just make some sort of a 'canopy' that I can flood with CO2, and use a cooled nozzle with the gantry to print direct to steel.