Factorize mul, div, weightedAvg to minimize bytecode size

[Rubilmax]

#66 (comment)

[MathisGD]

In favor of this.

Related to #57

[MerlinEgalite]

As long as it improves readability while preserving gas efficiency I'm ok with that

[MathisGD]

By the way, I don't think that it would change the bytecode size, but only the Solidity code size.

[Rubilmax]

I won't spend much time for now on this, given that I don't believe the compiler will be smart enough to compile to something as gas-efficient as the current implementation (because we need to abstract a scale input to abstract multiplications, divisions, which is currently a library constant and would be input as a function argument on which operations are performed - operations that I don't believe the compiler will recognize as constant operations)

I believe it requires ethereum/solidity#891 to be addressed first

[MathisGD]

I didn't know that it wasn't possible to import a lib in a lib. So we are stuck with this...

[MerlinEgalite]

Should we close @Rubilmax ?

[MathisGD]

I would like to see morpho-utils completely revamped to do general fixed point maths instead of Aave/Compound specific maths. In this case this could be implemented.

Edit: with this naming 👌

[Rubilmax]

This is just a random open-source idea that I think would benefit someone in the universe, but is blocked by a Solidity technical limit
Given it's an OS repo, we don't need to have everything sorted out here, we can just let ideas free and see if they are picked up, supported by others