I'm not so sure it works exactly like that as far as doubling the attenuation.
I've never looked at the results before from doubled identical notch filters so I can't vouch for the below but I believe it to be true coming from Dale.
From PAR's article about filters,
Scanner Filter FAQ | PAR Electronics | Filters for the commercial 2 way market, MATV, FM broadcast, laboratory, marine industry, amateur radio, scanner and short wave listening enthusiasts
"Can identical filters be cascaded to achieve a deeper notch?
Yes, with a qualifier. If the filters are directly cascaded, the resulting increase in notch depth will not be additive.
That is, two –35dB filters directly cascaded may only increase notch depth by 6dB or so. The proper method of accomplishing this would be to connect the filters together with a coaxial jumper whose length is ¼ of an electrical wavelength.
This takes into account the velocity factor of the coaxial cable."