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.
Yes, you can simply put a resistor in series with a ceramic capacitor. The lower the better from the point of view of bypassing, so I would aim at 0.5 to 1 ohm. If you have lots of space, the electrolytic is fine (in fact you can parallel the two), and they are cheap. There are low ESR electrolytics and ones that are not-so-low, read the datasheet. If no datasheet, no buy.
You should be able to read the numbers on the datasheets even if some of it is in a foreign language. If you're going to the market and picking shiny parts off of vendors displays without looking at datasheets you will get bitten. I've always been able to get answers to questions such as the load capacitance of a crystal without being the most amazing linguist around.
The x1117-3.3 is extremely cheap and very available in China so I don't see any reason not to use it. If you don't need the power dissipation there are better choices in SOT-23.
Best Answer
If you're trying to make a R-C filter, then your deliberate R should be much larger than the ESR (equivalent series resistance) of the capacitor else you will hit other effects that will mess up your circuit anyway. Yes, in theory, you add the ESR to your external resistance as in your example. But if this actually matters, then you're too close to the limit. Your example is good in that it shows the ESR is well below the noise level. You've got much much more slop in other areas than represented by the 1/2 Ohm added to the external 1 kΩ.
Take a look at any good capacitor datasheet and you will see that every capacitor only works properly up to some frequency limit. For small surface mount ceramics, this is usually at a few 100 MHz. Often this will be shown as impedance graphs, where the capacitor impedance magnitude is shown as a function of frequency. For the ideal capacitor, this would be inversely proportional to frequency forever. For real capacitors, there is a low impedance limit, then the impedance starts to rise again as the frequency increases.
There are all kinds of effects factored into the impedance graph. These include particulars of the dielectric, unavoidable parasitic inductance, and probably only in a limited sense ESR. Remember the "equivalent" in ESR. Most of it is not a real series resistance due to the construction of the cap, but a simplified way to presenting a host of other effects, particularly details that go on in the dielectric.
In short, something as simple as a single ESR number doesn't hold anymore when you get near the minimum impedance frequency and beyond, or the self-resonance frequency. If you stay far enough away from those, then ESR will be noise to a R-C filter. Conversely, if you find that the little bit of ESR would actually make a significant difference, then it's a solid clue you are running the cap in a regime where it's not really just a capacitor anymore. Remember that even good caps are ±10%, so a ESR that is 1% of the deliberate external resistance had better not matter, else you've got a tolerance problem in your circuit anyway.
There are two common places ESR does matter, neither of which having much to do with R-C filters. The first is to effect stability of a linear regulator when the cap is accross its output. Old LDOs were designed assuming there would be a electrolytic or perhaps tantalum cap on the output. These can be counted on to have some finite ESR. This ESR was considered in compenstating the control loop in the regulator. Without it, some regulators become unstable. More modern LDOs are designed assuming ceramic caps on the output, which have very low ESR. These regulators are specifically designed to work with a output capacitance down to 0 ESR. That is the only type you can safely put a ceramic cap on the output, since you can't generally count on them having some minimum guaranteed ESR. The datasheets generally only guarantee the maximum ESR, and that is quite low.
The second place is when suddenly dumping large pulses of current onto a cap, as happens in many switching power supplies. The current times the ESR represents a momentary apparent rise in the cap voltage, which often must be carefully considered.