Change all SubClassOf where RHS is an intesection to multiple axioms

[cmungall]

@turbomam and I are looking at some oddities in OLS

EBISPOT/ols4#619

This is what hail looks like in Protege:

'hail' SubClassOf
'object aggregate'
 and ('has member' some 'hail stone')

We should never have A subClassOf (B and C). This is always equivalent to writing

• A SubClassOf B
• A SubClassOf C

(this holds whether C is named or an expression)

While the two are equivalent, the latter is always easier for humans, and many tools make assumptions about structural form

(of course, I also think we should get rid of object aggregate altogether but that is for another day...)

[cmungall]

This also holds for depth of water ice - but I think this should be ECA