That discrepancy is actually easy to explain.
For better or for worse the eastern European power tubes are generally built with all the qualities of farm equipment: Mechanically rugged, works even in the face of large mechanical tolerances, huge spread in real world performance.
The Russians couldn't be bothered to make the internal mechanical structure of their power tubes to terribly high standards. Their philosophy was to build mechanically rugged tubes, inexpensively and relatively quickly, and then design their equipment to compensate for the resulting spread in the individual electric parameters of each tube.
In a beam power tetrode like yours, the screen grid current is hugely dependent on how mechanically well aligned the wires of the two grids are. In principle the screen grid is supposed to be located in the 'shadow' of the control grid, as seen from the cathode (this is the 'beam' part of beam tetrode). Doing this hugely reduces the intercepted electron flow of the screen grid, and thus screen grid dissipation. So for a given power level the American tube grids require less metal, resulting in lower capacitances.
In an American (Eimac) tube this mechanical alignment is pretty much spot on, to a tiny fraction of a mm. However, this comes at a cost of increased overall manufacturing cost, as the whole structure has to be built to exacting standards, to make the tube sufficiently mechanically rugged.
The Russians designed their tubes to not care about this too much. For that reason their screen grids are designed to dissipate higher power, compared to a similar American model, which is what you are noticing. The downside is the huge spread in all parameters depending on the exact grid and cathode geometry, for instance capacitances and overall transconductance.
It is the same reason why the Americans specify 0 (zero) dissipation on the control grid for similar tubes (4CXwhatever). Their control grids are gold plated to reduce secondary emission, yet that means the grids needs to be run at zero power at all times, as the gold plating can tolerate almost no heating. The Russians didn't bother with the gold plated control grids, so they can take a bit of heat. And you frequently need it.
However, whether the full screen grid dissipation is a good limit to set is a bit hard to say, without access to the constant current curves. You are running the tube at nearly half the maximally allowed screen grid voltage, so it may be that the screen current is quite a bit below the maximally allowed at your chosen operating point. Screen current close to the thermal limit may signal you have another problem, which requires a shutdown. It might be a good place to start though, if you have nothing else to guide you.
I would try to find other people's experiences with your tube, and check what the limits are on their actual operating parameters, including during tube-up. Alternatively, try to dig out the constant current curves.