Yes, keep your new crystal (52.5 MHz) in place from now on. For the first test, you want the setup you already were using at some point before, with the counter in high band, looking at the IF output, where it previously showed the LO frequency, and with the LO running from +15 V, which was the limit of your external supply, as I understand. This should show something around 2095 MHz on the counter, as before. The PLL should be running open loop without any IF feedback, which is what you want. Then you want to confirm that the PLL tune output voltage can rise and stay at somewhere around +17 V, which is about where I think it should go when open-loop. But, if the search oscillator kicks in, then it will need to be disabled. If all goes well, the tune voltage will go up and stay near 17 V.
Presuming the first test works as planned, the next step is to remove the external supply, and reconnect the internal tune voltage via P446, so now the LO is running near 17 V, and should be at a higher frequency, which should show on the counter. That tells you how high it can go without additional modifications, and it's hopefully well over the 2095 MHz result at 15 V. If it is, then there's a very good chance that when you close the PLL loop, it will run the LO up to the desired 2095 MHz and lock, and the tune voltage should end up somewhere near 15 V (where you got about 2095 MHz before), and be stable.
The last test finally closes the loop by restoring everything to normal - putting the 2LO in from the 7L13, connecting the IF out to the PLL, and enabling the search oscillator. Now you're looking for a stable tune voltage around that 15 V, which would show that the PLL is in control, and the LO should be locked at 2095 MHz. The counter isn't necessary, but you can measure the IF if you want, with the scope probe and counter setup as you did before, tapping into the IF amplifier signal, but be sure to not interrupt the IF to the PLL.The IF should be at the new 105 MHz, or at 52.5 MHz after the divide by two circuit. Now everything should be running about right, if the 2LO BPF isn't so far out of adjustment that there's not enough signal to get a good IF level.
In reality, you could skip the first test, but it's good to start from a stable point you know, where you had it set up and measuring the LO frequency. Also, if the tune voltage driver does not operate as I expect, then that issue would need to be addressed before the next steps can work..Good luck.