When I am doing complicated things in Rhino which are drawn from 2D, such as a group of outlines that become a solid or extruded plate, I often leave these on a separate layer, layer with the name “Work” after it. I often make these layers blue, but that’s a choice.
From this, you can extrude and perform other operations, and then hide the work layer.
When doing a series of Boolean operations on a solid, I usually put these objects on a layer ending with the word “Tool” . These, I often make orange.
With both curves to be extruded as well as many objects to be subtracted from others, these can be grouped together.
If you are used to a stack-based system, you can also make sublayers for critical and intermediate solids.
Yes, there are some detriments to a non-parameter non-stack based system, but there are a lot of advantages.
Working with Rhino can be very interactive and organic, in that many objects can be altered without looking up their parent. You are not limited to one construction plane, and your plane of reference can be anywhere and any direction you want.
Though this power requires organization. There is nothing in Rhino to prevent you from leaving curve and point piles lingering about. Yet, as you sort and layer your drawings and designs, you will likely arrange them into groups of ideas that mean something to you–instead of a list of circles that make…a plate.
As @Diego Krause stated, many of Rhino’s non parametric tools can be history-enabled, which allows the resultant object to be updated.
[I once made a machine animation in Bongo, Rhino’s animation model. In that animation, I was able to animate bending hoses by animating their parent curve nodes. Without the animation module, you can just as easy change curve points which affect either a pipe or a rat’s tail, if you will]