Actually, there is another possible consideration....or two:
When updating a DOL clustered index, the datarow the index key points to doesn't move as it does in APL. This can distort datapage cluster ratios if it happens a lot.
When doing large deletes on DOL, the rows are only marked for deletion. The housekeeper GC is the process that comes around and rewrites the page or deallocates as necessary. Unfortunately, each engine in ASE only has a HKGC pending queue of 50000 entries.....so if you overrun the HKGC queue (HKGC overflow in monOpenObjectActivity) then the housekeeper doesn't have a record of the overflowed records to delete. Same thing with a crash or shutdown with nowait. This is likely to leave the dpcr as if the deletes haven't happened yet - which could distort things a tad....or more than a tad if whacking huge amounts of data.