[r-t] Anything Goes vs Peals Mean Something

[Richard Smith]

I replied to MBD:

>> For instance, two extents of Minor and a 3600 of Triples
>> could produce this. You're not saying you'd reject that?
>
> No, of course not.

But isn't that exactly the point?  I don't believe you can 
properly distinguish this case from (certain special) false 
5040s of triples.  I fundamentally don't see how you can 
produce a rule that allows 'nicely' fragmented extents of 
minor interspersed through a 3600 of triples, while 
disallowing an 'obviously' false 5040 of triples that happen 
to be describable as a 3600 of triples with two 
'unnaturally' fragmented extents of minor interspersed 
through out it.  That difference is a value judgement and 
has no place in our new decisions.

RAS