No offense but you clearly dont know how a computer works, and on top of that, how pci-e works. first of all, u commonly hear pci-e lanes. thats because in the standard, EACH lane has a dedicated 250Mb/s (500Mbits/s full duplex (upstream and downstream combined)). A pci-e X1 has 1 lane, a pci X4 slot has 4 lanes, etc. So your 4X hd controller will have plenty of bandwidth for a whole buch of drives (in fact a total of 2Gb/s, which is enough to handle 13 individual SATA devices, with an extra 50Mb/s unused in the total allocated bandwidth for the 4x slot). Hope this answers your question.
on another subject, I am incredibly amuzed how nvidia implemented SLI. SLI takes the 16 anes for the first pci-e 16X slot and splits it over the 2 16x each with 8 pci-e lanes when sli is activated. now, how the hell does that allow the graphics cards to even work? well first of all, the truth is graphics gard dont need the full 16x bandwidth, in fact, they don't need the current 8X agp. AGP 4X will keep any of the g-cards happy. the reason for the extra bandwidth (that none of them use) is memory access. however, you can logically deduct that with the upto 256MB of on card ram, they really dont need access to the main memory. the bandwidth is actually used for texture/pixel loading. its not a bad thing to "waste" all the bandwidth, its jsut unnecessary and if 16 ever does become needed for the most powerful g-card ever, then u'll have the bandwidth. Cool huh?