if we assume that the electro-active species don't perform any chemical bonds with the electrode, than the main difference is the distance of the species from the electrode. For the species in the electrolyte, the cyclic voltammogram will show diffusion limited reaction while the species absorbed on the electrode will be limited by the electron transfer rate.
In general depending on the electrode, redox couple and ionic strength the oxidation/reduction for a species adsorbed at the electrode will be independent of diffusion concentration overpotential.and ohmic drop. As a result the oxidation and reduction peak separation (as measured in mV) will tend to zero, or in any case will be significantly smaller than for the same species in solution..
if we assume that the electro-active species don't perform any chemical bonds with the electrode, than the main difference is the distance of the species from the electrode. For the species in the electrolyte, the cyclic voltammogram will show diffusion limited reaction while the species absorbed on the electrode will be limited by the electron transfer rate.
In addition the above answers I recommend you to take CVs at different scan rates, to find out the surface confinement of the electroactive species under study. up on ploting the graph for E(potential) vs v(scan rate) ,you should get a straight line passing through the origin. if it is diffusion controled the straight line you got will be for E vs square root of the the scane rate.
As you described, preadsorbed Vs. species suspended in the electrolyte, if the preabsorbed species mean surface-confined, then supposedly, current should be linearly proportion to scan rates. Otherwise, for diffusion-controlled electron transfer, current is proportional to square root of the scan rate.
However, when the scan rate (e.g. > 100 v/s) is faster than the electron transfer rate, even for surface-confined species, diffusion-controlled electron transfer can also play in a role.