Heinz,
The question of more smaller devices versus fewer (or one) bigger devices really depends on too many factors to definitively answer here. There are so many layers possible to address when talking about I/O performance. There are dependencies here on your expected transaction workload, number of engines, OS platform, ASE version, underlying filesystem/raw devices, SAN configuration, ASE cache configurations/wash sizes, etc, and etc.
I _think_ the most straighforward answer I would give is that it's probably a matter of "best practice". That is, the approach I would _start_ with, is the thinking that "more smaller ASE devices, in general, aren't a bad thing".
There are considerations to be made on how ASE processes IO. The ASE configurations '"i/o batch size", "disk i/o structures", and "max async io's per engine/server" all play in to how many concurrent IO's can be issued by ASE, how ASE batches IO requests, how housekeeper checkpoints work, etc. There are pseudo device level commands such as "dbcc tune(deviochar,<devno>,<val>)" that assist housekeeper task to checkpoint faster, etc. Lot's of complicated interactions here.
If you have a very low transaction workload (including tempdb io), obviously this isn't much of a concern.
There are OS/Volume manager settings that control IO operations to volumes (veritas vxm settings in your case), driver/controller settings, and likely SAN settings to consider as well. I'll assume you are using raw devices and not filesystems because that adds even more to consider.
So, from my point of view without knowing anything about your environment, my gut always tells me, more smaller (obviously within reason) is better than fewer (or one) bigger because ASE doesn't really "know" that each ASE device is on the same LUN/Volume etc on the backend. And especially so if those devices are fast (ssds, or high SAN cache hit ratio, etc), and if those devices are write heavy loads. So, any "per device" limitations can be spread out over more control structures. In your case, it may not matter at all, but i'm pointing out a mindset that you might take going forward on the question in general.
Do yourself a favor and read the "Managing Workloads with ASE" doc at
http://scn.sap.com/docs/DOC-35088 and particularly the section that addresses IO considerations.