You didn't read my post, did you? Your first scaling is not linear, in fact it is highly non-linear.
However the numbers I have provided (you can find those by going through all data sheets. Also those numbers were also given by Ed Schempp at some time) show that the charpy-C notch toughness is approximately 5 times larger for S7 than for 440C (we are just talking about order of magnitude, there is no point in calculating that to the last digit). Column for the toughness of S7 is approximately 5 times larger than the one for 440C. Hence, the scaling must be pretty much linear and there can not be a significant offset.
However the numbers I have provided (you can find those by going through all data sheets. Also those numbers were also given by Ed Schempp at some time) show that the charpy-C notch toughness is approximately 5 times larger for S7 than for 440C (we are just talking about order of magnitude, there is no point in calculating that to the last digit). Column for the toughness of S7 is approximately 5 times larger than the one for 440C. Hence, the scaling must be pretty much linear and there can not be a significant offset.