Desktops are not as power limited as laptops and are not assumed to be needed to work in this configuration.
Doing this kind of active switching will require some kind of hardware support for the feature and means that your dedicated GPU output would have to be fed back to your iGPU or you have to have a lot of complicated multiplexing of outputs going on.
From the Nvidia Optimus Whitepaper the image below shows how the system is wired up.
This means that you would not be using the display outputs from your graphics card and instead you would have to render on the dGPU, pass the rendered image data to the iGPU and then display it. Not using the 3 or more outputs on the graphics card would limit you to however many outputs are on your motherboard. Having only one display would be a deal breaker for many people.
Otherwise motherboard manufacturers would have to integrate 3 or more display outputs that until this point have always been on the graphics card. This would complicate the matter or wiring and routing the motherboard.
For the high end gamers the extra delay of rendering then transferring data to the iGPU for display rather than the graphics card immediately displaying it itself could be unacceptable.
There was apparently a Desktop Optimus solution slated for release back in 2011, but never made it to market. Presumably this was due to lack of demand and lack of support from manufacturers who saw it as unnecessary when the power efficiency is less of a concern.
As a result both the iGPU and dGPU are considered to be completely independent on desktop systems and have very little of the integration that is seen on laptops.