I prefer the second method because for time.

Method 2 or 1.
I'd probably do method 1, but it depends on what the model is used for.

Method two is "easy", but results in a more complex shape that contains all the features. If a change is needed (i.e. remove the chamfer, or the groove at the top), it is more difficult to make those changes in method 2 (compared to simply deleting/surpressing the features in method 1.

It is less of a factor, but revolved geometry is generally inferior to analytical geometry. Analytical geometry comes from features like the extrusions used in method 1. Non-analytical geometry comes from revolves, lofts, sweeps, and other features.

While I'd likely use the method 1 approach. I would not use a sweep feature for the 3rd step. A revolve makes more sense there.

The 3rd method makes no sense. There is no reason to create bodies, then use a boolean operation to subtract one shape from another in that example. Combine is a great tool, but not for shapes like the ones shown.