No, that is irrelevant.

Assembly code mnemonics have a direct one-to-one correspondence to the actual binary machine opcodes; therefore, if something cannot be expressed in Assembly code mnemonics, it cannot be expressed in the actual binary machine opcodes.

Conversely, anything that can be expressed in the actual binary machine opcodes must also be expressible in Assembly code mnemonics - or, as Erik put it, "ANYTHING a given processor can handle is, by definition, possible in assembly."

High-Level Languages do not magically invent new machine opcodes; they simply save the programmer the tedium of having to express everything in the fine detail of machine opcodes.

High-Level Languages make many things practical that are simply not practical to write by hand in assembler - but that is very different to saying that they are not possible in assembler!