Many cheaper LCD panels have native color depth of 6 bits per channel. This makes manufacturers of monitors implement dithering to avoid banding. But what actually limits the panels in their color depth? Isn't it simply a matter of using 8-bit DACs instead of 6-bit ones? Are the latter significantly cheaper or what?
Inside the TFT panel driver, yes, the difference is just in using 6-bit DAC vs 8-bit DAC. In most panels, however, this is not insignificant because you actually need one DAC per column of pixels. So when you multiply the small required chip area difference by the number of colums, it starts to become relevant.
Note: some panels use sample-and-hold technology, so the number of DACs can be reduced, but then you need the hold circuitry for each column, so it is a compromise that may not be worth it for smaller (cheaper) panels.
Now, it certainly doesn't make a huge price difference, considering that there are a lot of other parts in a TFT panel. But it is still cheaper to manufacture.
Moreover, it also makes 6 signals less to handle. So another reason why 6-bit is so widely used for cheap panels is that the interconnection between the controller board and the TFT panel can be made simpler. This interconnection is typically made with LVDS signalling (serial communication), and for 24-bit RGB, you need a clock and four LVDS lanes for the data (five pairs total), whereas for 18-bit RGB, you only need three LVDS lanes for the data (four pairs total).
So it makes a difference, especially if the distance between the controller and the panel is a bit longer than usual (think about cars displays, or even just laptops: the screen is far away from the graphic card, and the wires have to go through the hinge).
So if the cheap products will be limited to 6-bit/color because of the interconnect constraints anyway, the TFT panel manufacturer better have slightly cheaper 6-bit panels in their catalog.