I would NOT use this product until it had been benched for at least 500 hours on a working proc.
Processor cores are affixed with solder, it any gallium spills onto it, then the gallium will corrode, and or dissolve (depending on how much gallium) the solder, which could disconnect the core.
Gallium is also very corrosive, especially to copper. With such small amounts of gallium I would think that eventually the gallium would just make a solid alloy with the copper, so that there would no longer be any liquid gallium left to fill in the peaks, and vallies between the core, and HS.
There should be a thin layer of SiO2 (glass) on top of the proc core, gallium wets glass extremely well, and would be a superior heat conductor. The major issue is the corrosion of the copper (or aluminum), and the proximity to the special solder that is used to affix the core.
It probably works great for the first few hours, or so, but after a while I would think that it will corrode the HS (whether copper, or aluminum), and then would be as effective as running a copper HS with no thermal paste at all.
Back when a few of the members (including myself) were fantasizing about building a liquid metal cooler, I thought of an open core cooler, where the gallium would be in direct contact with the core of the proc, as it flowed through the cooler. I decided that covering the sides of the core with RTV silicone would be the best way to protect it. Gallium will attack most metals, but not many rubbers, or plastics, and it shouldn't attack silicone. I never tested it to be sure though.