I think over-training is used as a sloppy definition for a variety of results from working-out. When I was running track seriously, I was often over-training, particularly from high intensity sprint work combined with weights over a period of weeks; but I would bounce back after 4 days relative rest in better shape than before. This is over-reaching rather than over-training; and it would happen far less when "on". The only problems would be bacterial / viral that everyone faces.
Cardio, though, I agree, can really mess you up - if done to a high level. I started doing 40 miles per week (granted, few on the board will be looking / needing that much) a few years ago, with 2 x weights per week, because my priorities changed. I've been fucked now for two years: dizziness, fatigue, depression, weight changes, flickering eyelids, constant swelling of glands, etc.
I was taking minimal supplements at the time - glutamine, vits, etc; so, I suspect I could have added some test in there to buffer the raised cortisol. But the problem with combining weights and high cardio is that, neurologically, you stress both the sympathetic and parasympathetic nervous systems. The parasympathetic is less benefitted from supplementary test, so if you over stress the system (which works antagonistically to the sympathetic), you raise the risk of implicating the other in the overtrained state; and even by adding test in the overtrained state, it is difficult to recover. Only rest will help - and it can take years, believe me. One of the main problems is that when you try to re-introduce some anaerobic work, inc. weights, this will actually stress out not only the sym- but the parasympathetic NS, because you are overtrainined, and you go back to square one.
If you live a very stressful life (as I did), and are already hitting the sympathetic nervous system hard, you need to be really careful with the cardio, even when "on". Look out for changing in mood (in particular, depression and aggressiveness (more moody than 'pumped', if that makes sense, and lack of motivation). That's the best guide. I wouldn't rely on test to buffer the (potential) dangers - and I certainly would bear in mind that orals usually have the effect of raising cortisol, so as acting to compound problems that are building up.
Don't want to sound like an alarmist prick, but thought it worth using my extreme case study to point out danger of too much of a good thing...