I have yet to fire the beast up, but do not anticipate trouble. It is like new inside.
Eico did a nice job with that transmitter from both a design and build perspective. I think your analysis is on target. If you follow the wiring diagram, you will see a low value (40 pF?) fixed after the plate blocking cap that goes to ground. A similar cap goes from 6146 signal grid to ground, putting them in series. I suspect that the low Q Pi allows you to get away with fixed caps, obviating the need for variable caps and adjustment with final changes.
I hope to pair the 720 with my 2B and Johnson T/R switch. Yes, I have a low pass filter.
There is a 140pF variable cap (C16) after the DC blocking cap (C15) but that is for plate tuning. C50 is additional capacitance for loading up on 80 meters.
Mike (MB) is also correct as in fact, page 2 of the 720 manual states this very thing, as C13 (68pF) is the "de-harmonicizer."
The Hallicrafters HT-40 and the Allied Knight T-60 both used a Pi-net grid driving circuit as well, but their grid shunt values were on the order of 9-10pF, because these sweep tubes have an input capacitance of about 23 pF. I have never seen any parasitic oscillations in either the HT-40 or the T-60.
I would recommend adding a 2.5mH choke (of at least 160mA) between the SO-239 center conductor and ground in case C15 shorts so that no HV DC gets into the tuner or antenna.
Phil - AC0OB