We had a pretty good discussion on this topic last week.
http://www.practicalmachinist.com/v...ex-vs-okuma-multus-vs-doosan-puma-smx-333656/
Understand that the more complicated the machine becomes, the more fragile it becomes as well.
Personally, I think a Y-axis, live-tool lathe makes a lot of sense for the right parts. However, I'm not a huge fan of sub-spindles
(on live-tool lathes, or B-axis mill-turns), except in a few specific cases.
If you have parts that have milled/drilled features on both sides of the part where the radial orientation is absolutely critical, then the sub-spindle makes sense.
But for general work, I don't think the sub-spindle makes a lot of sense, as it makes your setups much more complicated, and overall cycle times longer. And you almost always have one spindle just sitting there doing nothing... If you have a job that's going to run unattended, then the longer cycle times won't kill you much. But for general work, I think you could move more parts through the machine, by running all the parts thru the OP10 setup, then flipping the parts & setup and running them all thru the OP20...
If you have a twin-spindle, twin turret machine, then that's a completely different animal. Two spindles always in cut - that's a very strong case to try and punch holes in.
Citizen is also doing some really cool things with their single turret, twin-spindle machines, using their superimposed-turning function. They kind of blur the lines between a single turret machine, and a real twin spindle, twin turret machine. It might be the ticket for your 2" & under parts?
Miyano BNA-42 MSY - YouTube
But if you're going to be doing lots of compound-angle features, then you almost have to go with a B-axis machine. I would not consider them for *heavy* milling though - if a 2" indexable cutter - 3" face mill max - won't do it, then I wouldn't try it in one of these machines...