Conses are different in newlisp, but it certainly does have linked lists:
> In newLISP, there are no dotted pairs. Instead, each newLISP cell contains one object and a pointer to another object if the cell is part of a list.
> different evaluation model
That one is true, yes.
> no GC
No, but it implements McCarthy's original intent through the ORO scheme. NewLisp's ORO thing is very annoying if you ask me, but it does its job; it achieves the same thing that McCarthy mentioned in his paper: "This process [...] is more convenient for the programmer than a system in which he has to keep track of and erase unwanted lists. Its efficiency depends upon not coming close to exhausting the available memory with accessible lists etc."
Let's not delude ourselves here. NewLisp is not necessarily one of the "better" Lisp implementations, but it has S-expressions, Scheme-like function application and automatic memory management. I think that's reason enough to call it SomethingLisp.