By request here are a few distortion tests of some of the tweeters I have on the shelf. A guy from Finland asked if I'd seen John Krutke's ribbon tests and would I please comment on this. Yes, I often visit ZaphAudio for Krutke's tests of drivers; very useful.
If ribbons are used properly, i.e. high slope filters and not taken too low, I think they may deliver some excellent highs. As can be seen below, distortion levels go low from simple filtering; so low, I think it's irrelevant. My worst experience with ribbon tweeters is the Red Rose Music Rosebud mini-monitor. From the German Image HiFi magazine test it can be seen that this speaker has a 2nd order filter to the Aurum Cantus 130F1 midbass and - hold on - first order filter to the AC G2-variant ribbon. This speaker was almost ripping off my ears. Hard to believe this was designed by Mr. Levinson.
Anyway, it's always interesting if measurements can be be reproduced over time and location and with different instrumentation, so here we go.

Red = SPL/2.8V/0.25 m (for most tweeters this will produce 89-92 dB SPL at 1 meter. Actual level is important here).
Blue = 2nd
Green = 3rd
yellow = 4th
purple = 5th
Distortion lines raised 15 dB.
All tweeters having 33 uF in series, the ribbons must have, thus all others had the same regardless of impedance.
All tweeters were flush-mounted.
All graphs having 1/12 octave smoothing.

In conclusion I have to say that there is no correllation between distortion levels and which tweeters I like and which I dislike. So don't discard a certain tweeter based solely on distortion levels; it may very much depend on how it is used and which middriver or midbass it has to mate.

Aurum Cantus G2Si. Yes, high 3rd harm. dist. below 4 kHz. 2nd harm. dist. rising towards 10 kHz. This looks a bit strange, so what about the other sample?

Aurum Cantus G2Si, sample #2. Yes, 2nd harm. dist. at 10 kHz is the same as for sample #1.

Fountek JP2 ribbon tweeter. This is large ribbon and it seems like 3rd harm. dist. pattern is repeating itself. No rise in 2nd harm. dist. towards 10 kHz as seen for the AC ribbons.

Fountek JP3, my favorite ribbon tweeter. Same picture again. Quite low 2nd towards 10 kHz. Is this why I favour the JP3 over the G2Si?

Fountek JP3. Here with 3.3 uF in series. Let's take a look at the 3 kHz point. Here we have 3rd harm. dist. more than 40 dB down (remember dist. lines are raised 15 dB). -40 dB should account for 1% distortion, so really not too bad from a simple 1st order filter.

Fountek JP3 again, here with 3.3 uF/0.18 mH (simple 2nd order filter). I consider these distortion levels for insignificant.

Fountek NeoCD3. Very much the same picture as for the "naked" alu ribbon. I had expected these laminated ribbons to perform differently. Looks very much like the JP3.

SEAS 27TDC. Not bad at all!

SEAS 27TFFC. 3rd harmonic well in line with 27TDC, however some 2nd harmonic distortion at 6-10 kHz. Is this why I like this tweeter so much? I really favour the 27TFFC over the 27TDC....

Vifa DX25TG. A noticeable rise in 2nd harm. at 6-7 kHz. Otherwise excellent performance.

The Vifa XT25TG really has low distortion levels. Is this why some people think this tweeter is boring? I like it a lot. Peculiar dip in 2nd harm. at 7.3 kHz.

My farourite large dome, the SS 3806/8200. Used in the SP38. Extremely low 3rd harm. This "tweeter" can be used really low.

Another of my favourites, the SS 9800. I don't think it comes much better than this.

Not my favourite tweeter, the SS 8513, used in the 2.5 clone. I wouldn't take this tweeter below 3 kHz and I'd use a 3rd order filter. I think he SS 18W/8535 and the SS 8513 is an unfortunate combination. The 8535 cannot play upper mid/lower treble and the 8513 shouldn't be taken too low. Bad mix Mr. Tyler.

Morel MDT30, a tweeter I love to hate. I had a friend coming in with a speakers fitted with this tweeter and we could never make it work, so he left it behind in favour of the Vifa XT25TG. I really don't know why it was doing so bad. Apparently distortion wasn't the reason.

I like this Neo3PDR planar a lot and it was used in the Point75 construction and I have plans to use it some more. Here with the crummy plastic rear chamber as supplied. Still performing well.

Neo3PDR planar with an open back (dipole) as in the Point75. SPL is lowered slightly from removing the rear chamber, so is 2nd and 3rd harm. dist.

JBL LE20 tweeter. This tweeter has some significant higher distortion levels compared to most modern tweeters. 2nd harm at 5-8 kHz is really high.

JBL LE26. This tweeter should probably not be used below 3 kHz. 2nd harm. fairly low above 3 kHz. Some 3rd harm. at 4-5 kHz.

Wharfedale Super 3. This is really vintage stuff and it's doing much better than expected. Some 2nd towards 10 kHz, but 3rd is fairly low and all harm. goes flat down to 1 kHz.
No rear chamber used during measurements, so this is dipole.

Peerless 62CT pp-cone. Not really vintage, but I like them so much I wanted to see how the would perform here. Not bad at all.