Mathematically speaking, if we look at the outer reaches of Roman Mauretania (in particular, the city of Sala Colonia), we find that, from Rome, costs and time are comparable to traveling to Alexandria or Antioch, and much closer/cheaper than London (less than half the time, little more than half the cost).
So, we can estimate that, with the proper wind, Rome could get to the Canary Islands easier than Britain, and possible down to modern Mauritania, maybe even Senegal for a comparable investment to Britain. That doesn't really change the value of getting there much at all, but, given the Roman troubles in subduing the whole of Britain, you can pretty much count that as the outer theoretical limits of Roman power projection along the African coast. This doesn't take into account that sailing along the Spanish and French coasts on your way to Britain probably has many more decent ports of call at which to supply your ships than the Saharan coast.
I don't know as much about the Saharan camel trade, but quick googling indicates that it didn't really take off until the 3rd century, and it was after the Islamic conquests that it solidified into a reliable trade network.
As far as the slave trade is concerned, I don't know that the Romans would be all that interested in sailing around Africa in order to buy slaves. On the one hand, if they wanted to buy foreign slaves, they had plenty of Germans, Sarmatians, Arabs, etc. to enslave if they really needed more slaves, rather than take the long way around.
Further, the Romans seemed to prefer to get slaves as spoils of war, rather than just going out and buying them. Maybe its a holdover from their martial character, but it seems that, whenever there was extended peace (or, at least, an extended period between successful external wars), Rome tended to have a shortage of slaves. Suffice to say, its possible that the Romans wouldn't get too many West African slaves unless they bothered to conquer West Africa.