This has been a good discussion. I tend to approach the subject slightly differently when I teach this aspect of photography. I look at it more from the standpoint of the trade-offs one makes and I generally start with the discussion of what correct exposure is.
Sanjib's question and Bill's response partially miss what the intent of the exposure is. If the intent is to shoot an image like the one Bill shows in #16, then I would have given a similar response and probably would have gotten into writing about adding a fill light from a flash as an option. He does allude to this by saying "The metering - the question is "what to expose for?" That dictated by what you want." If the intent was a silhouette, the metering / exposure would have been quite different.
Ultimately, I suggest that what metering is all about is getting quality data to create the image you are trying to create. The answer will be different for someone who is looking to get a strong straight-out-of-camera JPEG versus someone with good post-processing skills who is working with raw data.
When it comes to the Exposure Triangle I tend to start with what the trade-offs that need to be consider are:
1. Shutter Speed - controls motion; intentional and otherwise. If freezing motion is critical is critical, then there is one answer, if a softer look of water in a fountain or water fall is important, then there will be a different answer.
2. Aperture - controls depth of field and the optical quality of the image. I generally am concerned with depth of field but some people get into a real twist when they are not shooting using the "sweet spot" of their lens. Wide open and some of the optical deficiencies of the lens design come out and stop down too much and diffraction softening reduces image quality.
3. ISO - As others have pointed out the base ISO is going to give the highest image quality (maximum dynamic range, maximum colour depth and minimum digital noise). All of that really doesn't matter if the shutter speed is too slow and unintended subject or camera movement blur have a negative impact on the image or if the image is not sharp enough where required because the depth of field was too shallow.
My usual work flow is to decide if I am trying to control motion or depth of field. I will make those decisions and then go with as low an ISO setting as I can (I generally do not use auto-ISO), for a given shutter speed and aperture. Obviously the focal length I am shooting at is also a consideration when assessing the settings I am using. I also tend to use the same preset white balance for the entire shoot, rather than using auto-WB, even though I am shooting raw + JPEG.
I generally use evaluative metering and on occasion pop over to either spot metering or high-light weighted and will always look at my histograms on the first shot of a set to ensure I am getting good data.
I do use exposure compensation or switch to manual mode in tricky light situations were I know the recommended exposure will not give the quality data I need for PP work.