[old_tv_nut]

The 100 pF does not need to be in the calculation because it's in series with the grid capacitance. 100 pf is so much larger than the grid capacitance that the series combination is only a tiny bit smaller than the grid capacitance itself.

C53 is already very small, plus you will always have the grid capacitance loading the coil, so adjusting C53 smaller for a higher value coil will not work. However, if you get a lower value coil, you could try increasing C53 to tune to 4.5 Mhz if necessary.

If you can find an aluminum core that fits the Miller coil, you might try that to reduce the inductance, but beware that it will also reduce the Q, possibly too much.

Any chance of removing some windings from the Miller coil without breaking it? If so, you could measure the resonant frequency in circuit before removing windings to get an estimate of how much to remove, and then proceed cautiously.