It has been widely discussed if meteor scatter double hop propagation is possible at all. A paper published at http://www.scit.wlv.ac.uk/vhfc/prop indicates that expected signal levels are much too low to be detected.
While the basic radar equation used in that paper is correct, there scattering area parameter is not used properly. The authors discuss the probability of very large scattering areas, but miss the fundamental difference between physical cross section and scattering area.
The scattering area used in the radar equation is calculated from physical cross section and wave length:
S= 4*p i * (PCS / lambda)2
where PCS is the physical cross section and lambda is the wavelength.
Now let's see what difference this makes. Let's assume a physical cross section of 104 m2. With our correct definition of S, this physical cross section gives a radar cross section of 7*107 m2 - that's in the order of what we need for a successful transmission!
Does this proof that double hop meteor scatter propagation is possible? No, but it shows that large scattering area can be obtained from objects with moderate physical cross section.
The full calculaction can be found here (PDF print of the MathCAD calculation sheet) created by Volker Mühlhaus, DL5DAW.