You make a very good point here. Sredni's "path dependency" is less and less noticeable the more turns on the secondary. I did the experiment with a three turn secondary and posted the results in this forum. Sredni almost calculated the voltage between nodes A and D, VAD but suddenly realized that if he did the calculation he would show that KVL works perfectly, and backed up quickly...
You clearly have a very fuzzy recollection of those events.
I computed all the values in your silly circuits in less than 15 minutes and got the results right. Including the voltage along a diameter in the case of perfectly circular and concentric geometry.
You still cannot understand that voltage IS path dependent.
Cyriel Mabilde in Youtube did a similar experiment too with a five turn secondary and also demonstrated that KVL works perfectly. Now I am waiting for some insight from Sredni of what would happen if we replace the resistors in the loop with capacitors, but he is MIA...
Mabilde, from the depth of his garage, is another KVLer who cannot imagine a path dependent quantity. And this is an old movie that is being rerun over and over. The KVLers propose their 'killing' experiments that should make us "Armchair Nobel prize physicists" fly away to another galaxy. Then we post the solutions according to classical ED (it's not 'our' theory, it's plain old classical electrodynamics) and you fade to silence for a while, except coming back with muddy recollections of events.
It happened with the 'two secondaries is series', it happened with the straight partial coil, it happened with the multiturn coil, it happened with that sentence by Belcher (Jesse is still touting it in his boilerplate answer on his channel and he is forced to ban users who do not agree with him to make them 'fly away to another galaxy')...
...and now we are at the ring with two capacitors. As if these capacitors could change something.
So, here, are the results for the following values of capacitors
C1 = 4.7 uF , C2 = 22 uF
freq = 50 Hz
emf = 374 mV
I get - from simulation and without even invoking MEAS, just by eyeballing the plots
VcapL = 308.5 mV, VcapH = 65.5 mV
Guess what I measure with a true RMS multimeter?
VcapL = 308 mV, VcapH = 66 mV
And nothing, ok almost nothing, in the copper joining the caps.
So, what are the revolutionary results that you said were bad news for 'team Lewin'?