well ... not so fast
1 byte = 8x bits
there is no problems to call it either bytecode or bitcode, in general people will use "byte" because historically you used 8 bits to encode a single character of text in a computer, but both terms are valid.
Still there are differences, in a VM (Virtual Machine to execute code), you will use bytecode if your instructions are stack-oriented, whereas in a VM with bitcode is not. Which means instead of loading values into registers, a VM using bytecode loads values onto a stack and computes values form there.
So why Apple use the term bitcode then?
it's because they use LLVM (Apple is a big contributor to LLVM) and LLVM bitcode is closer to machine-level code and isn't bound to a particular architecture which allow LLVM to use an arbitrary number of logical registers.
Other VM, like JVM (Java Virtual Machine) and AVM2 (ActionScript Virtual Machine 2) on the other end use bytecode because they use a stack-based approach, they POP and PUSH data to/from the stack.
See Stack based vs Register based Virtual Machine Architecture, and the Dalvik VM.