Do you really think the cat is really dead?

I just re-read most of Tony Leggett's review article, Testing the limits of quantum mechanics: motivation, state of play, prospects. He considers that different interpretations of quantum mechanics broadly fall into three classes.

1. The statistical interpretration.
This claims that quantum state amplitudes (i.e. wavefunctions) have no reality but are merely a calculational device to calculate the probabilities of the outcome of measurements. Questions about whether a cat is dead or alive before it is measured are
ruled to be not "meaningful." The most "trenchant" advocate is Leslie Ballentine.

2. The orthodox interpretation.
QM amplitudes are "real" at the microscopic level but effectively not real (or at least not relevant) at the macroscopic level. This is claimed to be because of decoherence one can "For All Practical Purposes" (FAPP) never observe quantum interference between macroscopically distinct states.

3. The many-worlds interpretation.

Quantum states do exist in "reality". But when I think I measure the cat is dead this is actually an "illusion" because the alive state is "equally real." I am just unaware of it.

Leggett's view is that 1. and 3. are largely hermeneutical gymnastics that allow one to avoid discussing the problem. His agenda is to propose experimental tests of 2.