The original question was asked based on the understanding that the film ISO was a firm setting and the adjustment to shutter speed and aperture are not, based on my limited understanding of film cameras.
Film speed is better viewed as a recommendation rather than something fixed. What it is is an expression of how a particular film will behave when exposed to a variety of scenes and then put through a particular development process. Unlike digital, film (negative film, here) doesn't have a linear response to ever-increasing brightness - the highlights tend to compress, meaning that, at a certain point, the density of the negative is no longer proportional to the amount of light that hits it. The highlights might still become more dense on the negative with an extra stop, but the change in density will be less than what you might get with a mid tone. This compression can make things tricky when you come to print or scan because the printing or scanning process has to compensate for the compression to correctly render the original differences in brightness in the various parts of the scene. At the other end of the scale, the shadows need sufficient exposure to effect the chemical change that occurs in the emulsion - too little and you just don't record the details (negative density is too low).
Film box speed is the maker's recommendation for how to expose the film and goes hand in hand with a recommended processing regime. The idea is that the film will record densities that tend to stay away from the compressed region in the highlights while still getting sufficient detail in the shadows.
Both things can be varied - you can rate the film at a different speed and over or under expose as a result, and you can adjust the development process to compensate (more development, or less development, or use techniques to develop some areas of the negative more than others). For some variations, there are documented recommendations (like pushing film - under-exposing, and then developing it more), either from the makers or photographers in the field. However, varying these can result in poorer negatives (eg, pushing tends to increase grain and lose shadow detail - but it might be your only option for getting a shot).
If you're just getting started with film, just expose at box speed and see what happens. There's little point in diving into the complexities of exposure (and processing) until you're in a position to make an informed choice. Do box speed first, and maybe try a few shots of the same scene at different ISOs to see how they compare with box speed. From that, you'll begin to understand what sort of difference it can make.
Of course with digital auto ISO is a wonderful thing and something I use almost exclusively. I'm far less confident with film other than using a metered system such as with my OM2, I've never shot a meterless camera before and so it's all a bit of a learning curve!
If your camera meter works, then use it - they're generally perfectly fine. If you want to try shooting without a meter at all, read up on the sunny 16 method.
Regarding that blog article, it seems badly written or plain mistaken in places to me. This bit...
"Because color negative film usually gives the most pleasant results when overexposed, a lot of film photographers rate their film at half box speed (ISO 200 instead of ISO 400) and expose for the shadows, which results in 2-3 stops of overexposure."
...doesn't seem right. As I understand it, exposing for the shadows means taking a reading from the shadows and then adjusting the exposure to place the shadows at a certain number of stops away from the mid tone (meters always give a reading that they think is correct for a mid tone - so-called 18% grey). In other words, if you meter a shadow area and want it to come out suitably dark in the final result, you need to under expose from that reading. If you exposed at the shadow reading and used a standard processing method, your shadows would render as a mid tone - which ain't right, generally speaking. So far as I'm aware, 'a lot of photographers' (I wonder how many, and does he have their names and addresses, or is he just making an unfounded assumption?) don't meter in the way he describes.
His invocation of the zone system...
"An incident light meter always shows a reading for neutral grey, which is zone “V”. Instead of zone “V” you assign zone “II” to “IV” by literally holding your meter into the shadow (the darkest part of the picture)."
...seems a bit odd. Last time I read The Negative, you assign the meter's 'zone V reading' of the shadow to the zone that you want to place the shadows on. (The meter sees the shadow as a poorly lit mid tone - zone V - and wants you to give it extra exposure so that it renders as a mid tone in the final result.) Since you know it's really a shadow, you have to decide how dark it should be compared to a true mid tone in the scene (2 stops darker, 3 stops, or whatever). What he's saying doesn't seem to make much sense, and it's questionable to bring up the zone system in an article that is presumably aimed at beginning film photographers without covering the principles behind it properly (something that took Ansel Adams a lot more than half a paragraph to do). How many beginning film photographers have enough knowledge of the zone system to grasp what he's talking about? And do any that do have such a grasp need to read his article? Virtually none, and no they don't.
I would ignore that article. He might have a method that works for him, but I don't think he's describing it very well. I think some if it is likely to lead more to confusion than improved understanding.
So... Box speed with a meter. Do a few over and under exposures (take notes). Learn sunny 16. 90% or more of the time, your exposures will be fine. The ones that aren't are the source of learning more - post the iffy results here and ask questions.
