Comparing the OW tweeters is somewhat difficult. While there are subtle differences among them, the differences are extremely minimal. I suppose it would be like comparing two tube amps at $10k each. Sure, there would be minor differences, but they would both sound extremely good.
I will convey a few snippets regarding the OW3 (gold) and OW4 (platinum) dome tweeters. My explanation will be somewhat circuitous, but should suffice. These tweeters were ordained because folks in S.E. Asia really wanted gold and platinum COLOR domes. Following this growing need, Oskar consumed a few years in the workshop/lab refining the coating process in his domes so these colors could be obtained with no adverse effects. Oskar is very tedious in these regards. While the previous process was good, his latest process is... well... perfect. Other manufacturers simply coat the fabric, press the dome, and ... it's done. This is the complete antithesis of Oskar's methods. He is extremely tedious when hand coating his tweeters - using refined compounds in the places where needed. AND, oh by the way, he measures EVERY tweeter and culls those that don't measure within spec.
I am going to take a tangent and get on my soap box for a short period herein and convey something in-print that I have conveyed verbally on many occasions. Quality control in loudspeaker driver manufacture is an immense issue and can be extremely problematic. It has "bit" me on a few occasions, but my problems were VERY minor compared to the stories I have heard from others. It would appear to the consumer that loudspeaker drivers are all the same, and that they all should perform the same from on production run to the next. It would also appear that loudspeaker drivers all should perform the same within the same production run. In almost ALL cases, these are are very poor assumptions. And, in almost all cases, the results are audible by the educated listener. Most listeners, without time at the measurement bench, won't be able to isolate the problem to a specific frequency anomaly. The listener can simply sense that something is... hmmm... wrong. And, indeed there is often something... wrong.
However, there are VERY few documented cases of... the other guy's loudspeaker Q.C. problem rooted in poor driver Q.C. . This is simply because the guy revealing the other guy's loudspeaker Q.C. problem could result in a lawsuit for the conveyor of the problem. Almost nobody in the business can afford one of these. There is 1 guy, Dennis Murphy, who is essentially immune to this and has revised several crossovers to accommodate what I believe are driver changes across production runs.
http://murphyblaster.com/content.php?f=marble.html . I also suggest reading this article, and perusing the measurements too
http://www.stereomojo.com/Small%20Speaker%20Shootout%202007/SmallSpeakerShootout2007Part1.htm .
I would like to take a moment herein to convey a few solid praises for one of my brethren in the DIY hifi endeavor - Danny Richie. I wish to illustrate that good Q.C. is extremely important - even when using modest quality drivers. Danny Richie purchases huge quantities of modest quality drivers, then Q.C.'s the snot out of these drivers and only uses/sells those that Q.C. correctly and measure good. The primary reason for Danny Richie's success is that EVERYTHING he sells works well. The secondary reason for Danny's success is that he performs excellent design work. I have recommended Danny Richie to potential customers on many occasions and have NEVER read or heard of any negative customer feedback regarding Danny's products.

I am... going to stay on my soap box just a few more minutes regarding a related issue. It is my belief that marketing departments and consumers agree that there must a synergistic engineering magic necessary to create a good/great loudspeaker. Perhaps there is some degree of luck involved. There are a myriad of factors present in written material that are apparently very critical to the design of any respectable speaker. Candidly, this I find this rhetoric abhorrant.
In summary, it is my assertion that tedious quality control for loudspeaker drivers and completed loudspeakers is extremely important and sadly absent from loudspeaker rhetoric.
I should be VERY clear that I will not support my assertions with a plethora of data. I can not and will not risk a lawsuit. As such, you will have to trust these assertions based on my reputation. And, I am "from the government and here to help". aa
Okay, back to the correct topic...
In all of this, I will offer a summary of characteristics based on my own experience and the feedback I have received over the years. First, all of the OW tweeters sound VERY similar. I could live with any of them in my living room. The differences presented among them are profoundly minimal. However, the very minor differences lean towards the following suggestions.
1. If you are unsure which OW tweeter you want, choose the OW1. Most folks (@ 85%) in most settings have found this tweeter most pleasing to the ear.
2. If you need 90db sensitivity or prefer a slightly ribbon-like air/edge, choose the OW2.
3. If you want a fancy colored tweeter, choose the OW3 or OW4. Also, the recent revision in Oskar's coating process and fabric revision was intended to create a sonic signature more like the OW1.
4. If you want to cross the tweeter low (i.e. @1650hz 4th order), use the OW1FS (with ferrofluid). I crossed this tweeter to the SEAS W22 driver several years ago and the tweeter did NOT distort/suffer when pushed hard.
Hopefully these comments are helpful, and I apologize of my soap-box commentary was a distraction.
Sincerely,
Dave