Print speed is also a factor though this can depend on the geometry of a particular model. I wrote about this some time ago but still have not captured some images to reflect the impact of speed on a particular model. Some models are better at being printed fast than others. Some absolutely demand a slower speed. I assume this is all about the distance of moves being made but it also seems to matter where the moves happen. One thing that I know for sure is that if I want to bet on better quality as opposed to worse quality then I slow down the print speed!
This leaves the size of the nozzle. This will likely not be important to many people depending on how much of a hassle it is to change a nozzle. On my first printer, a Prusa I3, it was not too difficult, but still took some time. On my second printer, a dual extruder, Makerbot Clone, it was a little more of a hassle because of the two extruders. On my Ultimaker 2 it used to be huge hassle...until I upgraded to an Olson Block via 3DSolex. It is amazing. I can't quite do it in the 10 seconds they claim but I do get it done pretty quickly.
This has made printing with a smaller nozzle something that I can do with ease. It has allowed me to design some parts that would be unprintable with the stock .4mm nozzle (such as my Bailey Bridge kit).
Given my fondness for interchangeable nozzles I had also come to the conclusion, with no actual evidence to back it up, that a smaller nozzle noticeably improves print quality. This was based on my perception of things that I have printed but not with an actual side by side comparison. Below I will do a side by side comparison.
Many models that have been designed to be 3D Printed already acknowledge that extrusion width is a limiter and will allow for this by ensuring that details can be printed with a "standard" size of nozzle laying down a "standard" width of extrusion. If the designer is recreating a physical object in smaller scale this can help them by limiting how much detail they need to preserve. I think this was a factor in the detail present in the tanks designed by M_Bergman though I can not be sure of this. They print great with the standard nozzle (0.4mm, extrusion width 0.48mm) on my printers. On the other hand, if you look closely, you can see that they print just a little better using a smaller nozzle (0.25mm, extrusion width 0.3mm).
My test case is the top of the turret for a T-34-85. It has some good detail and allows for a nice slice of the top of the tank to minimize print times. This is a handy way of doing a test like this, namely using the slicer to print just a section of the model.
The below images illustrate that the smaller nozzle does have a positive impact on print quality even in a case where the model is well designed for a normal nozzle. What surprised me, however, was that the difference as pretty subtle as I expected more!
The standard 0.4mm nozzle is on the left and the smaller 0.25mm nozzle is on the right. Two features to point out with the first being the view ports at the base of the commanders hatch and the second being the hinge on the smaller hatch in the foreground. In the first case it is not evident (to me anyway) that the quality will be different though from the picture it clearly is better on the smaller nozzle. The hinge does look like it should be crisper from the smaller nozzle but in reality the difference is harder to spot!
It is easier to see the impact that a smaller nozzle can have when looking at an example of small details. The above model is the same turret shrunk by 62.5% (difference between 0.4mm and 0.25mm nozzles). In this case there is detail that will be printed showing on the slicing rendering on the right and no detail at all rendered on the left. The slicer ignores stuff that is too small to try and print!