No. Neither method is perfect for estimating the number of crimes. The difference primarily is in surveys and police.
The police know only a certain amount of crimes. Many crimes are not reported. For example, victims are embarrassed, victims are afraid of threats of retaliation, or victims think nobody cares about them so there's no point in reporting. The police know only a little bit. And the police may or may not write it down if they get it reported, depending on budget, time, and priorities. So this is probably a minimum estimate.
As to surveys, social scientists have made an expensive and energetic effort to track down people and interview them. It is suspected that this is a much easier way to report crimes, and thus crimes that really aren't serious or really didn't even exist are reported. So it is possible that this is an overestimate.
There are two different methods that social scientists use to estimate numbers, and that is the difference between those two columns in table 6. They're both based on the same estimate of cost. If you look at the column on the far right, where we have the costs per victim, those costs per victim are used for both columns in table 6.
Does that help?