I use Rhino for all my drawing now. I was playing around with switching with Rhino V4 (as it introduced layout space), but it wasn’t until V5 started to beta that I really committed to switching (a lot of layout bugs were fixed with V5).
It has been a long time since I have made a drawing in AutoCAD, so I don’t think I can make a very fair or complete comparison at this point (I vaguely remember solview and soldraw). The impetus for changing was more about I loved modeling in Rhino over AutoCAD, and trying to translate the model from Rhino to a drawing in AutoCAD was a real pain, and involved a lot of make2d, shuffling back and forth, and then having to do it all over again when something changed.
The main differences that I still recall are Rhino’s annotation tools are a little bit less customizable at the moment. For instance, there is no justification in dimensions, they are all left justified, so you you pull a dim then edit it to add text, and your text wraps, you can’t center it, it will justify left. Things like that. But the speed gained by doing it all in Rhino, and the ability to use the model as the drawing (not having to make2d it), more than make it worth my while to make the switch.
As for the LISPs, what did they do? It is possible there are scripts or other workarounds that do what you need.
For the actual making drawings in Rhino, there are generally two approaches. Either rely on the display to do the much of the drawing, or convert to curves and use that for the drawing. I mostly use the display approach, but in actuality it is some combination of both. Mostly display, then use some line work in layout space to emphasize, clarify, or add hatching.
We had a conversation about the differences between these two methods on the old newsgroup, and I’ll copy and paste my thoughts from there:
The display method certainly has some great advantages, it is fast, the display is very flexible (I can turn on shadows, colors, move my clipping plane about). As soon as I update my model, a large part of my detail is updated as well. Without all the make2D information everywhere, the model stays very clean.
It also has its disadvantages. With lots of details on the page all shaded, the layout page can get very sluggish to navigate (it would be nice if there were some viewport caching to help speed this up). You don’t have as intimate control of line weight and style. You can’t just select a couple of lines and reduce their weight or change their line type or remove them. It isn’t very good when you need to show a lot of hidden line info, I find it is easiest just to draw this in manually in paper space. I do mostly architectural stuff so this works fine for what I do, but if you need to show every single hidden line in the model, it falls short. There is the print quality issue as well. If you are relying on a shaded display mode to get the view you want, it is going to print raster (although vector paper space objects can still print vector), which can (sometimes greatly) reduce print quality.
I find there is also a bit less “fudge-ability” when using the display method. If I want it to show up in the model, I have to model it, and I have to model it well, where as if I am using the make2D method, I can put a place holder in there, and address the part just in 2D land. There are also places where the display method breaks down, everything is going along fine and then someone wants to change that elevation from projected to developed. Now you have spent that time modeling and annotating, only to have to do the make2D thing anyway because of a special condition (same thing for reflected ceiling plans).
Is the display method perfect? Absolutely not. But for the work I do, I find its advantages are a strong enough draw that I try and work around the short comings. It definitely wont work for everybody, but the way I look at it, the more I work this way, the more I complain about what I find and don’t like, the better and faster the display method will get, and the less I have to rely on make2D