By definition, a Cherry MX clone should be grown from the DNA of another Cherry MX switch … wait …Azteca wrote: ↑A Clone by definition must be identical in every way.
There is no official definition of "clone" in terms of technology. However, being identical is not part of any definition I've ever seen. I go with the principle that a clone serves as a drop-in replacement, i.e. if you replaced the original with a clone, it would still work. An IBM PC clone can be put on a desk in place of an IBM PC, and still run all the same software.
In terms of switches, this means that it needs the legs in the same positions, it needs to fit the same size hole in the mounting plate and expect the PCB to be the same distance from the plate, and it needs to take the same keycaps.
That's why I don't class Omron, SMK or KPT switches as Alps clones, as you can't use them in place of a real Alps switch. Kaihua PG1511 series however can be used as a direct replacement to Cherry MX switches in the same keyboard. Hence, Cherry MX clone. It's not as though the internals are different either: they're a straight-up copy of Cherry's design, with the expected simplifications. You can see a comparison here:
http://www.kailh.com/gb/Newsdetail.asp?Newsid=37
I don't consider the design of the contact mechanism to be part of the clone definition, but since Kaihua copied that too, it's really quite obviously a clone. It doesn't mean that PG1511 switches are inherently poor; after all, they're copying a reputable product.