It is not about emergency shutoff. That is not the function of any disconnect. Their function is to turn something off, and keep it off, while regularly planned maintenance is conducted. But what about maintenance on the disconnect itself? You can say that they simply need to turn off power upstream of the disconnect. But if the intended maintenance includes thermographic analysis, you need to do that with the disconnect energized and with its cover removed. And if the maintenance technician needs to lean over the desk, in order to remove the cover, he would be in an awkward physical position and there would be an increased risk of slipping.
This is not published anywhere, and is not official, but my belief is that a key reason for maintaining working space clear of all interferences is this: so that if the worker contacts live parts and gets shocked, and if the shock is severe enough to cause his hands to contract so tightly that he is not able to let go, then perhaps the only thing that could save his life is when his muscles lose all control and the weight of his falling body pulls his hand away from the panel. The owner of this forum, Mike Holt, tells a story of personal experience that is something along these lines. He says that if he had not fallen, he would have died that day. But consider this: if there is a desk surface underneath the disconnect, and if the worker is leaning over that surface to get to the disconnect, and if the worker does get a shock, his body will not have the freedom of motion that would cause him to fall onto the floor. Just think in terms of that desk taking away the workers one chance, however small that chance might be, but his one and only chance of surviving the shock.