In Haswell, 4.8w out of 22.1w for the core were used for the decoder for integer/ALU instructions[0]. According to this[1] analysis of the entire ubuntu repository, 89% of all instructions were composed of just 12 instructions (all integer/ALU).

From this we can infer that for most normal workloads, almost 22% of the Haswell core power was used in the decoder. As decoders have gotten wider and more complex in recent designs, I see no reason why this wouldn't be just as true for today's CPUs.

[0] https://www.usenix.org/system/files/conference/cooldc16/cool...

[1] https://oscarlab.github.io/papers/instrpop-systor19.pdf