The reason they are defined differently is based on the differing mechanisms:
All of the monomer in step-growth (of which condensation is a subset) polymerization are reactive, and they are consumed very quickly, forming (mostly) dimers which then go on to react further. At this point the average length of chains is 2-- not really a polymer, but also relatively monodisperse. However this reaction is going to continue, and you'll get long chains after some time. But the monomer itself is gone.
In addition polymerization (of which radical is a subset) there are very few reactive intermediates, and many monomers get added per chain. After 50% of the monomer is converted in addition polymerization, you have some really long chains, swimming around in the remaining monomer (half the original amount). This contrasts the other situation.
Basically, you can't measure monomer concentration to reasonably monitor the progress of a step-growth polymerization, while you can do such for addition polymerization. Conversely, you can generally monitor functional groups to monitor both kinds of polymerization.