Yep, that's the going price. Trackers are not cheap. However, neither is a similar size non-tracking pole mount: The trackers are about $6,000, while a similar non-tracker is about $2,000. So, adding a tracker would add $4,000 to the cost or about $1.33/watt.
Please note that the original question was about how to design a sun tracker and not about the financial viability of such a scheme. I ruled trackers out early in the discussion because trackers are ground mounted and therefore see less of the sky. That eliminates the major benefit of tracking, which is harvesting sunlight at low angles (morning and dusk). This low angle light is blocked by ground clutter (trees, buildings, hills). It might be done on a residential rooftop, but I don't think it will be practical unless one wants to ram an 8" pole through the roof of the house, and anchored into the foundation.
I think you might mean for an additional cost of $1/watt. Actually, it's a bit more because that's just for the additional panels and doesn't include the cost of the addition mounts and frames.
The current fashion in rooftop panel layout is to use micro-inverters and just scatter the panels all over the roof. This works well when the light is temporarily blocked by trees and buildings or when there's not enough roof area available. Some of the panels could be aimed to pickup low angle light (dawn/dusk) if there's nothing in the way. This reduces the peak power production, but increases average and total power production by providing some power at most any time of the day. It also allows the use of a mixed collection of different size and shape panels, which is aesthetically disgusting, but saves on panel costs.