Radial stiffness is almost entirely dependent on the radial stiffness of the rim (if steel spokes are used) which is why deep v rim are raidally stiff. If alloy or Ti spokes are used this is no longer true and the spokes do change radial stiffness.
High radial stifness allows few spokes to be used as the rim will deform less under load from lumps and bumps so the NDS spokes (as these are under less tension) will be less likely to lose tension. This helps spoke life.
This is a test has explained the benefits of 2:1 lacing with respect to lateral wheel stiffness but it does require a stiffer rim to pull it of as you do end up with fewer spokes unless you have a 27H drim and go for 18:9 lacing but not many rims around like this. For DS lacing on a 2:1 pattern why not use whatever you would use for that number of spokes DS. So for 16:8 then 3x DS would be sensible unless you build 32 pokes wheels 2x.
Wheel builders interlace mainly because it is what we have always done, we like the look and because it does no harm. Factory wheel builders to not because it does no harm and it make building the wheel simpler and that mean more money for them.
As for hub wind up as already said it is not a problem nor is wheel unwind up in general. For 28 spoke wheel I prefer 2x even though the spokes do not exit the hub in tangential way. Tortional stiffness will still be good and lateral stiffness is slightly increased. Also spoke do not cross the flange, I dislike that. Although 3x can work. You can worry about tortional stiffness more on disc brake wheels and then the way you lace will change a bit. Even than many 28 spoke disc brake MTB wheels will be 2x.
Thanks bm0p700f, to ask, with 2:1 lacing isn't the small DS bracing angles still an issue for DS lateral stiffness or has this been resolved in other ways (deeper/stiffer rims and larger DS flange to improve DS bracing angle, etc) ... thanks