Dear all,
Thanks for all your helpful comments.

I probably should have elaborated my conclusions at greater length. I will do so, which will respond to a number of the points that have been raised, and then start responding individually to the rest.

Bob's World's point on 'mini-me' armies is actually bound up with my statement that 'on the demand side, only way to build sustainable armies is
consensus through national dialogue.'
The only way that new armies in these countries can be sustained is if they're 'locally owned,' to use security sector reform jargon. ARVN in South Vietnam, built up as a replica of the ROKA to deter threats across the DMZ, was definitely not locally owned. Meanwhile, apparently the Ethiopians did significant reforms all by themselves and this created an army which the political class were happy to find resources for, thus be sustainable.
Pardon me as I repeat stuff many of you know, but I'm just trying to illustrate my argument.
The SANDF after 1994 is the security sector reform poster child of a locally owned army; I understand from contacts that despite all the doom and gloom we hear about the SA Army now, the core of the deploying infantry for PKOs in Africa is pretty good quality.

So yes, not a highly technical US-type army. Local militias, reserve forces only, maybe a small central force; not a 'mini-me.' In Sierra Leone and East Timor, both times, a reserve force has been considered but has been dropped. In Timor and in Kosovo, ideas have been floated for armies that do national development tasks as a primary role. This is because if, politically, we cannot disband the army, we need to give it something worthwhile to do, and avoid the common problem of lacking a role.

Infanteer's point on the PRUs is an extremely good example of a 'locally owned army' that I would be advocating. Maybe the CIDGs would be also.

Bill Moore's point is also good. Maybe, a la Haiti, we do not need an army at all. My full theoretical model goes through an assessment and programme design phase where the actual requirements are identified.

'culturally appropriate and sustainable' = locally owned, in SSR parlance. This is what I was aiming at.

I came to the conclusion re consensus through national dialogue though talking to a very experienced South African ex-infanteer, which reflected some of the stuff in the OECD Handbook on SSR. It's what they tried to do in S Africa after the mid 1990s. It's been tried in Liberia but too late to influence the contractor programme, and was first done ahead of decisions in Kosovo.

"Why? I really think that is the million dollar question, how do we help develop security forces that will win?"
This is what I mean by 'strong armies.' History's answer via Tilly and Jeffrey Herbst is let state boundaries move - darwinist evolution of states.
For the SFA community now, if I may be bold enough to say so, we cannot in virtually any circumstances, expect to build armies that can win when we import Western models.

This is a bit of a long post, and I must go. I will respond to the rest a little later. Thank you all very much for your comments - they are much appreciated.

Kind regards from New Zealand,
Colin