My seat-of-the-pants understanding for load capacitors (corrections invited) goes like this:
When a crystal is cut for a certain load capacitance, it is measured with that capacitance across it during final factory trimming. There is nothing magical about the value. It is simply a way of saying, that if you design your circuit to present that same capacitance, then your crystal will be within the stated (.005% or whatever) tolerance.
So, you add up all the capacitance in your circuit, and then add in what's needed to bring it up to the spec. We'll use your numbers. The stray capacitance due to the traces on the board obviously will vary with the board, so let's guess 1.3 pf. A number I made up, to go with the capacitance of the microprocessor's oscillator, stated to be 1.7 pf. So, we've got 3 pf in parallel with the crystal. The crystal wants 18pf, so we have to make up the 15 pf difference with discrete parts.
Since the two load capacitors are in series (Gnd->cap->xtal->cap->Gnd), we double the cap value to 30pf. Two 30 pf caps in series give us the 15 pf we're looking for.
Note 1. I tried searching for typical PCB stray capacitance. It was all over the map. Suffice it to say, that as the hardware gets smaller, the capacitance will keep getting smaller. A lot of typical values claimed less than 1 pf.
Note 2. If there is more capacitance than spec, the crystal will oscillate at a lower frequency than specified. If there's less, then it's higher. You can see, that if you want to trim the oscillator to spec, it's easier to shoot for a lower capacitance and add some later, than to try the opposite.
Note 3. For fun, look up "gimmick capacitor".
Note 4. My "seat of the pants" explanation is sufficient as an introduction, and this technique works in many cases, but not everywhere. For a more in-depth look at the EE principles behind those capacitors, see this answer.
Your title asks about the value of the capacitors, and I think that has been adequately covered- you should match the nominal value to the specified load capacitance of the crystal (when in series with each other, and subtracting some allowance for input and stray capacitance).
The Q of a typical crystal resonator circuit is very high (maybe 100,000), and a small change in load capacitance won't affect the oscillation frequency by much. The equivalent "motional" capacitance of the resonator is quite high, so the pull effect of the load is small (typically measured in ppm/pF). If you are not using the crystal for a time keeping clock, it probably won't make much difference for you- it will vary with the crystal and load capacitance, but, say 5pF might make 30ppm or 100ppm difference in the oscillator frequency.
Since the capacitor might be 22pF, 5pF is a lot of change, so the tolerance and temperature coefficient is not very important. It's also cheap and easy to find almost perfect capacitors in the capacitance range used for load capacitors- ceramic NP0 types with tolerances of 5% are the cheapest and most available, and they're always rated for at least the voltage required (Vdd + 1.2V is certainly enough). Take the 27pF value- a Samsung CL10C270JB8NCNC is 5% tolerance, 50V, maximum drift of +/-30ppm/°C** and insulation resistance in the 10G range. All for $7.54 for a reel of 4,000 pieces, Digikey price. The difference between microwave and ordinary NP0 caps would not be noticed at 16MHz (except, of course, for the much higher price of the former). There are all kinds of complications (voltage coefficient, high temperature coefficient, microphonics, aging) associated with high value ceramic capacitors that don't apply much, if at all, to NP0 parts.
TL;DR Bottom line- if you use the most common NP0 ceramic capacitors in your favorite size, your circuit performance will not be limited by the capacitors in virtually all cases.
** Note that a 30ppm/K change of the load capacitor would likely contribute less than 0.1ppm/K change to the oscillation frequency (the temperature changes will be dominated by the crystal itself).
Best Answer
Most of the manufacturers I use do have this data available (because if they don't, I won't use that manufacturer). It will be a "typical performance" curve and not a specification, but it will be published. It might not be available on the distributor website, but it will be on the manufacturer website.
ESR is not usually a critical parameter for a signal DC blocking application.
I usually take the value at the bottom of the dip in the |Z(f)| curve as the ESR. That will be good enough for most DC-blocking applications. If you really need to know the phase shift at some specific frequency, you might need to use a more complete model (but also, variation from part to part might make your careful modelling irrelevant).
It is even more important to consider the DC difference that will be across the capacitor in your circuit.
High Q and low ESR are just two ways to say the same thing.
Ultra-low ESR is not typically critical for a dc-blocking application. As long as you're using ceramic (NPO/C0G) parts, I wouldn't spend time sorting by ESR.
You want to look at the capacitance stability with bias voltage and temperature. For your application you should probably be using NPO/C0G parts.