In my case, I have found that estimates for the telephone number are more conservative than with address. I do not know if this is typical with G.fast since I only have made the comparison with my own tel. number and address. Interestingly enough since install, for telephone number the clean range lows and handback have dropped (by ~25 and ~50 Mbps respectively for downstream), but the impacted range lows and handback have improved by quite a bit (by ~40 Mbps for downstream).