Poured and Printed Nests/Fixtures
Poured Nests
Why pour a nest/fixture instead of machining it? Two reasons: one, a lot of parts, especially decorative parts, have such complex shapes in the interest of aesthetics that the CAD models are too huge for our desktop CAD programs or require five and six axis machining centers that can be unavailable to you or expensive to use. Second, a lot of those same parts are not allowed to be marred in any way or they're scrap.
Material
The overwhelming material of choice for a nest is urethane. Molders have special additives that control the shrinkage so that you get a nice .005-.007" [.13-.18mm] clearance around the part for loading/unloading ease. Choosing a hardness is a little different for a nest than a seal. In this case, the harder the nest, the easier it is to do something as simple as bolt it down and the longer it will last. But, the harder it is, the more likely it will be to mar the surface of the product, the very thing you're trying to avoid by molding. If your nest is too soft, it will deflect more, which may affect your machine function, and it starts to pick up debris that will scratch your parts, so it's all a balance. You will likely be in the 40-80 Shore A area for nests (see plastics post for explanation of the hardness scales). Just note, anything below 80 Shore (A) is very difficult if not impossible to machine, so they'll have to mold in holes, extra pockets, etc, if your application requires something softer.
Concerns/Limitations
Realistically, the tolerance of true position you will end up with of your part is around +/- .06-.10" [1.5-2.5mm] and can also be off +/- 2-5degrees in pitch, roll, and yaw because the molder is literally pouring around a part (ok, some are better than this, but plan for worst case). You will need to account for all that variation in the rest of your design, which isn't a big deal if it's a nest for a manual operation, but could require slots and compliance/teachability in any automation or even mechanization. Also, since it's poured around an actual part, that part defines the max part size that the nest will accept, so you either need to have the biggest part as your pattern (hard to determine that on a limited sample size and complicated geometry) or you can have them wax (I think it's wax) the part to make it bigger but still have the same shape. Finally, no two nests will be exactly the same, so you have to expect/plan for dealing with the differences if you have multiples or for in the field replacements down the road.
A few Design Tips
First, it's best to relieve the nest for everything except the bare minimum that will contain the part. This helps in several areas: A large portion of the cost of a nest is material, so reduce material, reduce cost; less surface area to mar the part; easier load and unload of the part; and if it's multiple pieces, you can replace only damaged or worn portions of the nest. Here are a couple of examples of this idea:
Another thing to worry about is literally as simple as mounting the fixture, especially in durometers below 80 Shore (A)/ 30 Shore (D) (see plastics post for explanation of that). Above that level of hardness, you can machine in bolt holes, locator holes, clearances, etc, after molding and the material is hard enough to torque mounting bolts down to hold it in place during use. But below that, the material has so much "give" that it's virtually impossible to machine and when mounting it into the rest of your design, you'll either deform the nest, or the bolts won't stay tight over time. The way around that is through inserts, either ones you design specifically, or commercially available items. This could come in the form of metal sleeves slightly shorter than the thickness of the nest which give the head of a bolt something solid to tighten down on, or thread inserts that would allow the nest to be bolted down from below. See pic for examples. The best case is to find something that has a retention feature that will hold the insert into the plastic over time. The molder will be able to hold the relationship from insert to insert pretty well (not dowel pin well, but let's say +/- .01" [.3mm], but just remember that the relationship from inserts to the part itself is subject to the kinds of tolerances discussed above, so something somewhere in the stackup may need adjustability.
The final design tip is about color. Urethane can be poured with a wide range of coloration, so don't be afraid to make use of that when dealing with change out nests or if it's the background to a vision camera and you need to provide contrast.
Printed Nests/Fixtures
Since most of you reading this are probably a generation younger than me, I'm not going to pretend to know more than you do about a technology that you've been raised with, but 3D printing/SLA of nests is a very real solution in today's world. They can print in a variety of materials of different durometers, colors, and physical properties. The reasons you may choose this over poured nests are: they can be printed with a 3D model of the part with mathematical offsets for clearances rather than using a part with imperfections/tolerance issues in the molding process; the nests can have very complex, precise features (tolerances on good 3D printers are down to less than +/- .005" [.13mm] now), even including printed moving parts (although it's technically possible to print the moving part right into the overall nest, I would recommend separate parts held together with conventional hardware because you want to be able to replace the weakest link); I believe you can even mix materials in the same printing, so you may be able to do softer durometer for the part holding bits and harder durometer for the mounting/exterior surface which then allows all of the mounting holes, and even possibly slightly oversized dowel holes to be printed into the nest with close tolerances to the part holding features; and turnaround may be faster depending on the situation. The reasons you may want to favor poured nests are: possibly the 3D geometry is not available from your customer, so you have no choice but to use a physical part; if multiples of the same nest are required, pouring may be cheaper; and in general, even on one nest, you have to be careful of the costs. 3D printing is always hailed as a very cheap manufacturing alternative, but I've seen details cost $1,500 whereas the equivalent poured nest is probably $600-800. As in most design decisions, it will come down to a cost/benefit/risk analysis as to the best way to go.