Right the gpu was probably idle. It was a cpu load to test full system running, aka every part needed for a functioning pc since a cpu can't run by itself, and was tested just loading the cpu. It was also tested using a power meter at the plug for just the tower, not the monitor and tower. Hardware meters are more accurate than software which attempts to read from motherboard sensors.
Another way they test in some power consumption benchmarks is to use an amp meter clamped over the wires leading the cpu power plug from the psu. Software reporting tools are mediocre most of the time and don't necessarily test everything properly or take everything into account.
When comparing to an intel system, the same/similar hardware is used. Obviously the motherboard cannot be the same since they require different sockets, but it too uses a motherboard, ram, data drive, gpu etc same as the amd system. Both tests for amd/intel in these comparisons use a fully functioning complete tower so it's a fair comparison pitting cpu/system to cpu/system.
It really depends on the personal outlook of the user as to whether it's an issue or not. It may also depend on someone's region where energy cost has to be factored in. For instance the cost of electric in England runs around double per kwh what it does where I live in the U.S. so it might be more important to someone living there.
Heat is another factor, amd's 8 core cpu's produce more total heat volume from power consumption than intel's quad core cpu's do. Partly because they're less efficient per core, partly because they're running twice the processing cores. Often times the amd 8 core cpu's get a further bad wrap because when it comes to performance in applications they compete closely head to head with intel's 4 core cpu's. (Close enough for comparison purposes). When it takes twice the hardware (cpu cores) to achieve the same performance it's another aspect of the inefficiency.
It doesn't mean the fx 9370 is totally incapable, it just means it consumes more power and generates more heat to do the same task(s). Whether or not that's an issue depends on the individual. For 1 home use pc it may not matter much. If someone's energy rates are higher it could matter more to them. If it's a business with 50-100 pc's it could matter much more.