Why we choose constant $T$ and $P$ for the definition of partial molar quantity?

Why we choose constant $T$ and $P$ for the definition of partial molar quantity?

It might be not a physics question, rather, a philosophic question though.

For a state function B, partial molar quantity of B is

$$
frac{partial{B}}{partial{n_{i}}}_{P,T,n_{j}}
$$

I understand the usefulness of this definition. With this definition, it is possible to connect the chemical potential with the partial molar Gibbs energy.

$$
dG = -SdT + VdP +sum{{mu}dn_i}
$$

therefore,

$$
mu=frac{partial{G}}{partial{n_{i}}}_{P,T,n_{j}}
$$

But besides such usefulness, it is possible to select any other constraints (like S and T) to describe the change of a state by adding the component $i$ into the system like,

$$
frac{partial{B}}{partial{n_{i}}}_{S,T,n_{j}}
$$

I mean, it is just a specification about the condition.

Is there any justification exclusively saying ‘partial molar quantity’ is $frac{partial{}}{partial{n_{i}}}_{P,T,n_{j}}$ ?