Well Nick...
All I have to say is everything that is written about Brain May's equipment lists his phaser as a Foxx Studio model 7.
I spent quite some time with multiple board images verifying the components.
Lowering the value of the two capacitors that you mentioned (C16 & C21) will only cut back on the bass response a bit. (less apparent phasing) The .1µf value has been verified.
R34 is 6.8M on every Foxx Model 7 and Regal RE-35 phaser that I have images of.
If you are hearing strong phasing without the signal being affected, I would look elsewhere for an issue.
If you recall, we added C23 (.1µf) to the feedback path.
Try lifting one side of C23...see if it makes a difference.
I was planning to do just that when I get around to building this thing.
Did you use a 300K for R13 in the LFO?
The 330K could make a difference in the sweep.

