If I can give a simple rule for simple 35mm photographers that would be:

If the subject is not in contrasted light (not a high SBR subject), put the dome in front of the subject pointed at the camera and you'll be fine;

If the subject is a contrasted one - half in shade, half in full sun, imagine the fašade of a building not entirely in the sunlight - then your choice is simple: with slide film place the dome in the sunlight, with negative film place it in the shade. Point to the camera and you'll be fine;

Only exceptions to the rule above are the cases when one has important details in the top highlights or important details in the bottom shadows. In the former case close a bit (let's say half a stop) in the latter case open a bit (half a stop).

The exceptions are due to the fact that incident metering will end up placing a very bright and strongly lit subject high in the film curve which, for slides, means a region with small detail, not much texture (when not burned). By the same token, incident metering will end up placing dark details in the shade in a region with small detail, not much texture (when not blocked).
A studio picture of some flour or sugar (something perfectly white and very, very reflective) when using incident reading would correctly render the flour, or sugar, perfectly white but that means not much texture if using slide film. Basically no slide film has good texture for perfectly white subjects. By closing we lose "pure whiteness" but we get "texture". Our mind "adjusts" for pure white because it knows sugar is white.

The normal case in real life is snow in the sun. With slides just use incident metering and close half a stop. The snow will bend toward "dirty snow" but the texture on the surface will clearly describe it as snow. Your "mind" will then bring it to pure white even when it isn't.

Incident metering is good for us because there is no need to place anything, to figure how grey is the subject. When we use reflected metering any metering cannot abstract from the question "what is the reflectivity of the subject". When we use spot metering we cannot abstract from the question "how is exactly shaped the characteristic curve of this film".

When we use incident metering in 99% of the case we have no thinking to do at all and it will yield good results regardless of subject reflectivity and film curve. Incident metering makes life much easier.

That's my experience but, again, I only use small formats and never deal with more sophisticated approaches such as ZS and BTZS.