In addition to the already mentioned issues (voltage drop from charge current in panel (and other) resistance, voltage reduction from hot panels, voltage drop in the diode), having extra cells doesn't hurt anything (except your pocket book B-) ).
Solar cells are essentially constant-current devices (if they have enough voltage to push the current). Each absorbed photon creates one electron-hole pair. Up to a certain fraction of them can be separated by the internal doping, with the electron fed to the negative terminal and a hole filled by an electron from the positive terminal (releasing a little energy as heat).
Hook a panel to a load at a lower voltage than its potential output and the panel's output will drop until the current is the amount produced by the least illuminated cell. The panel will heat up a bit more than if it were driving a higher voltage (because the electrons from the positive electrode have a bit more energy when they fall into the holes). But it would heat up still more if it were just sitting in the sun with no load and all the electron-hole pairs were all recombining without taking a trip through the external circuit and releasing the full 0.6V electron-volt of bandgap difference energy.
Similarly it's OK to short a panel: It heats up the same amount as if it were unhooked. The heat is distributed a little differently (some in the wiring rather than all in the semiconductor). But that's not an issue. As a result of this, some solar panel controllers avoid overcharging the battery by shorting the panel(s) upstream of the blocking diode, in order to maximize power delivered to the load. (A parallel, shorting, transistor wastes power when you don't want to use it because you have all you need. A series, circuit-opening, transistor wastes power due to its voltage drop when you want to use as much as you can get.)