Does this moa argument fly?

[John Bois <jbois@umd5.umd.edu>]

Moa and Elephant birds from Recent times were larger than any extant
birds.  And they lived on islands (NZ and Madagascar respectively).  Two
hypotheses have been offered for this phenomenon of gigantism on islands:

1. The absence of competition from large mammalian herbivores allowed them
to adapt into the large herbivore niche.

2. The absence of predation by mammalian carnivores allowed them to
evolve into large body size.

Predictably, I favor #2.  And while there is no direct evidence either
way, I wondered whether the following would be considered support, strong,
weak, or otherwise, for #2.

Moa and Elephant bird were both believed to be well under 1000 kg.
In Africa, large herbivores are limited by food availability while medium
sized herbivores (under 1000 kg.) are limited by predation (this from,
_Megaherbivores_ by R. Norman Owen-Smith).  The reason for this is, of 
course, that the smaller size of the medium herbivores affords predators a
better success rate.
Now, if you could stock islands like NZ with African 
herbivores and predators, it seems fair to say that moa would fall within
the size range that would be more likely to be limited by predation rather
than competition.
        And, I would argue that they would be even more severely limited
than mammalian herbivores because their juveniles should be within 
preyable size for longer than, say, ungulate young.

Is this too hypothetical to be of any value at all?
Comments?
Thanks in advance,
John Bois.