I understand that there is only one pole in the open-loop transfer function (which is G(s)) of type-1 PLL , therefore the use of final value theorem to obtain DC gain will cancel out the only pole (which is the denominator 's' within G(s))
The problem is in using ambiguous terms. "DC gain" in the quoted lecture is referring to the phase detector + loop filter, not the VCO transfer function respectively the product of both, the loop gain. The VCO pole doesn't cancel phase errors, you need a finite control voltage to maintain the operation frequency which corresponds to a respective phase error in a first order PLL.
The answer is actually : A type 1 PLL has only one integrator in the loop, that from the VCO , therefore the use of final value theorem to obtain DC gain resulted in finite value for Type-I PLL
Note: Screenshots below are from Floyd Gardner book.