To be honest, I'm not sure if it would have a marked impact on their current numbers. I would posit that most of their lines that have currently been activated were on existing phone lines, and possibly phone lines that individuals actively use. This is just a thumb suck, but I would say that probably 15-20% of current ADSL users would end up dropping landlines, but that's probably close to only 70,000 lines - certainly not a large percentage of the Telkom phone line base.
Where it *will* have an impact though is on new lines, you would see a greater decline purely because they're not making those numbers back on new activations. However, you will likely see increased activation of ADSL due to a lowered perceived barrier to entry for end users since those who predominantly use cell phones (an increasing proportion of the population) will be more comfortable with getting ADSL for home if they don't feel they have to buy a "phone line" that they will never use.
On a technical level, cable-based connectivity is still (or rather should be - ZA quality is very poor) the most stable and fastest option out there. While wireless and cellular companies are quick to claim multi-megabit connections, reaching those theoretical high levels are nigh on impossible due to any number of factors of which distance is a major one. ADSL is the better option if you want fast connectivity at home and office (and fibre is the same), so the value proposition is *there* as long as it's done correctly. It's just a pity that right now, in an effort to protect their existing business (a battle they will ULTIMATELY lose regardless), Telkom has hobbled the product in such fundamental ways.