Goal setting strategies : is the S.M.A.R.T model smart enough?

Here are two well-documented stated goals that can be considered S.M.A.R.T.  - one considered a great success the other a great failure, both coincidentally from the US.

An email thread I was on recently with thousands of smart recipients questioned the value of setting specific goals at a high level; that they can damage an organisation more than they help.   Are S.M.A.R.T. criteria enough?

The larger the organisation, typically the larger the number of explicitly different goals throughout the organisation at different levels, functional and regional groups, individuals, etc. - and the harder it is to align and set them so that they are a compelling, non-conflicting set.  That some group achieving one goal will help another group achieve one or more of their goals directly or indirectly - and definitely not disadvantage them.  This linking of goals is critical, but what kind of links?

In the S.M.A.R.T model the only stated element that directly addresses other goals (or activity supporting those goals) is the R = Relevant.  Relevance requires explicit relationship with other activity/goals.  The rest of the elements can all be calculated in a much more isolated manner (though nothing is actually isolated, e.g. Time normally has to be spent on other goals, but they need not defined explicitly).  This R citeria element is, I'm beginning to understand, the most critical 1 of the 5 - get this wrong and not only is that goal at risk, so are others they are linked to.

Unless there is a compelling argument otherwise, of all the elements in S.M.A.R.T. goals, the most important is R=Relevant.  And I'd better give that more focus that I have done before.  Also am thinking of seeing what can be done to make it possible to link goals together between different individual and teams in the most popular goal tracking tool we use at work.