That question cannot be answered. "Those Pins" and "specific Pins" is not specific enough. Some Pins will be passed on through and can be reused again and again. Power and ground for example. Perhaps A4 and A5 which CAN be used as analog Input/Output OR as an I2C Interface which, by design, can be shared.
So, to answer the question one must know what pin and what shield in order to determine whether that shield will gobble the pin or allow it to be further shared.
Look at the wiki link mentioned on the page you linked to. Six pins are used by your shield, and the rest pass through so you can use them of other purposes.
So the ones which do not pass through should simply be avoided? Is the reason they aren't closed off completely simply to allow stacking additional shields on top - or is there some other benefit to that design?
That shield uses +5, Gnd, A0, and D4-D10. D10 appears to be setup as an output.
So D0/D1 are available (but typically reserved for Serial comms) and
D2, D3, D11-13, A1-A5 are free.
So one might connect up some SPI devices to D11-13 with slave selects coming from the free pins.
Or maybe D2, D3 are used as interrupts to signal that a read of analog pins is needed.
Or whatever.
Having a screw shield under the LCD shield could make solid connections to the pins easier without having male pins coming out of the top of the LCD shield and blocking the view of the screen. http://www.crossroadsfencing.com/BobuinoRev17/
Maybe have an RTC circuit in the prototype area or something.
Ah- so the main advantage to not totally blocking D4-D10 on the top would be to allow creative routing via a screw shield or something like that... ok cool, thanks for the help!