Let's assume the secondary is open-circuit so that it has no effect on the the primary. When you apply 450V across the primary, that doesn't mean the primary winding has a voltage of magnitude 450V through it because total EMF = 450V - primary = rate of change of magnetic flux = Lp*d(Ip)/dt. With static I, total EMF = 450 - primary = 0, and so primary = 450V which, as you pointed out, would give an infinite current for a 0R primary.
Where there is a changing magnetic field, the electric field isn't conservative and so you can't assume that components in parallel will have the same volage through them. Suppose you have a ring of 1R in a uniform magnetic field normal to the plane of the ring. Suppose it changes at a uniform rate so that it induces an electric field in the ring with the total EMF around the the ring being 1V. The voltage across a 0.1R segment = 0.1V is in parallel with the rest of the 0.9R = 0.9V segment, yet the voltage drops are not the same. In a conservative field where total EMF around ring = 0V, you can say V_0.1R = V_0.9R.
Some people recommend this paper. If you want I'll email it to you.
R.H. Romer: "What do 'voltmeters' measure? Faraday's law in a multiply connected region", Am. J. Phys., 50, 12 (1982), pp. 1089-93