Well it makes sense for unreported rapes to be committed by someone close to the victim. Such things are obviously tricky to estimate, but Swedish authorities estimate it at 20% which was supposedly confirmed in a 2001 study of the extent of violence against women, funded by the Government of Sweden and the Crime Victim Compensation and Support Authority.
Anyway it would make little sense for a super-liberal country would have poorer report rate than other countries especially with their already huge rape rate. Even if all these unreported rapes were done by someone close to the victim(let's assume a native swede, family member/friend) it would still leave small number of immigrants committing majority of rapes, at least using figure from 1996-
77% * 80%(reported rapes) = 61,6%.