I think you might find this to be a useful resource https://reprap.org/wiki/MetalicaRap .
Its a project to create a metal sintering 3d-printer using electron beam melting. There's a lot of detail on the electron gun and its focusing system.

But what is causing the different phasing as you say? We use two type of E-Cap and only one shows the obvious larger ripple.

I meant the relative phase between the switching of your MOS rectifiers and the signal you are rectifying. For example, slow turn off would result in current flowing back to the pick-up coil, and therefore higher ripple in your caps.

As I said - try it with schottky diodes and see what happens.

Also your 25V MLCC caps running at 15V will be at about 40% of rated value.   Murata have a useful tool for this stuff https://ds.murata.co.jp/simsurfing/index.html?lcid=en-us

Generating xrays requires accelerating electrons to high speeds then slamming them into a metal target.  The neon gas fill in a nixie would likely stop the electrons ever reaching speeds where significant xray generation would occur.

But thats just my opinion.

N₂ emission at 337nm and 357nm is classed as UVA, so what you really need is sunscreen.  Or long sleeves. UVA causes skin damage.
Its worth noting that the UV in the arc is a lot brighter than it appears as your eye sensitivity drops off rapidly below 400nm.  The 337nm and 357nm peaks are 3-4 times brighter than the 400nm  you can see, and your eye is roughly 1/10th as sensitive to 400nm as it is to green light around 550nm.

Anyone own a light meter than covers the UV end of the spectrum?

What a beautiful thing! I was puzzled by the oval electrode at the bottom, then I realised its a puddle of mercury 
Looking at the plasma might not be too healthy for your eyes, depending on what kind of glass was used.

The wikipedia article on ignitrons would suggest a reasonably high current ignition pulse is required - probably similar to driving a big thyristor.
 
Oh look - a nice datasheet!
https://frank.pocnet.net/sheets/119/n/NL1001.pdf
So the trigger appears to need 200 volts, and a peak current to at least 30A.   To me this looks like cap charged to about 300V with a series inductor to control peak current and a thyristor switch. And a diode across the ignitron trigger to prevent reverse volts.

Doesn't handle as much current as I thought it would for something that big.

At those prices you'd be lucky to find any silicon in them at all. Got to be a scam.

I had one, and it blew the IGBT and main diode bridge. So don't go hoping to salvage those parts, thats probably why the microwaves are dead.

Paraffin wax will shrink as it cools and may crack.  You can make it tougher and slightly flexible by dissolving hot-glue sticks in the wax.

Nothing wrong with using a 555. As you have found you can get better control over the output than the usual self-oscillating 2 switch converter. The only real advantage of the 2 switch ZVS is simplicity.

Is there a reason you don't want to use a triggered spark gap? Or maybe a hydrogen thyratron?