You need to consider the capacitance and the on resistance independently (because their origins are distinct)
1. For junction capacitance, you need to consider two different cases (1) when forward biased and (2) under reverse bias. Why?
(1) when forward biased, the junction potential decreases, flow of majority carriers increases and the thickness of the diffusion layer increases. The diffusion layer is almost exponential shape (extends upto infinity) but we can use an effective thickness.
(2) under reverse bias, the junction potential increases, the thickness of the diffusion layer decreases and we have practically no majority carriers.
The current contribution from the minority carriers stay about the same in both cases.
The capacitance is formed due to majority carriers: high doping means thinner junctions.
Standard spice model of a diode has a const capacitor in parallel.
It is not difficult to measure the junction capacitance of a diode with simple tools.
Yes, most of the resistance for a forward biased diode is present in the junction (which is void of majority carriers formed by recombination of two different types of majority carriers).
I hope I am clear.