I am actually more of an engineer than a scientist but I do like to investigate and solve problems and have plenty of time and tools to do it. Since I was able to build a 3D printer from scratch and solve all the problems I found on the way, despite never having seen one, I am sure I could do the same with a laser cutter.
My original plan was to buy a K40 and hack that but a GF seems like a much better starting point although a lot less hackable due to the secret sauce in the cloud.
Obviously I could be completely off base, but it seems likely to me that the splash at any given speed/power has a natural scale/spatial freqency (same order of magnitude as the kerf). Meanwhile, the engrave has whatever spatial frequency corresponds to the lines per mm. Interesting things may happen when you get beats between those two frequencies. Which would give you something that had a Y-axis repeat.
Meanwhile, Iām also wondering whether this has to do with the way the GF is powered. As I understand it (possibly mistaken @dan please chime in if you care to) the GF has a beam that is on continuously, modulated to whatever power the cloud software is currently telling it to produce. Many other laser cutters have beams that are pulsed (per inch or per second or whatever) with each pulse adjustable by controller, and gaps between pulses. If thatās accurate it could also make interesting differences.
Iām running a test of my TaiChi logo coasters and your settings are causing definitely different results than what I used.
I engraved with power of 1% and speed of 335 in/min figuring that slower and higher power would mean deeper engraves (which in the 1/8" might be an issue with carving away too much.
So your 20%/200in per min would seem to suggest a much deeper and more āburning/meltingā type of engrave. It is deeper by a fair amount (Iāll have to wait until itās finished to measure). The lines/ripples/striations in it are a bit better than I got at my settings but now Iām seeing the crosshatch pattern you mentioned. Canāt tell yet if the ridges will wipe away like yours but Iām going to do another test with even higher power & lower speeds to see if I start to zero in on results more like I was expecting.
Will report back in a bit.
Wouldnāt ripples make peopleās star gates look more realistic though? 
Great! 
Iām still thinking that we need a lot more practice/experience to dial these things in - at least I know I do. I havenāt even begun to test variances in the Engraving, because Iāve just had too many other things that required immediate attention.
And I kind of hesitate to get too deeply involved with settings right nowā¦because I know theyāll be changing.
I got that fuzz layer that washed off you mentioned. But I still get more & deeper lines/grooves than I see on my other machines. It was a deeper engrave (by 0.045") than my initial attempts at 1/335.
Iāve got a 20/100 in there now.
Since these havenāt been dialed back yet, Iām not terribly surprised. Wait until they float the update and see if it still does that.
I was thinking of what @dan said about āsplashā, and if that has a serious impact, since these machines are really juiced right nowā¦it might be making the splash worse than they will ultimately be.
I really donāt have a problem with it thoughā¦itās not much of a pattern - just a slight texture.
(And we havenāt tried de-focusing the beam any either.)
The ripples are still there and in the same places when the beam is defocused.
An interesting test would be a 2 by 2 grid of squares with slightly different sizes, engraved as separate operations. If the ripples line up across engraves done at different times and different shapes then they are linked to the motion and not due to some unstable splash effect.
Unfortunately, your further speculations will have to go unanswered for nowā¦Iām done with testing for this. (Got a few too many other things to do at the moment.) 
(But maybe you can answer the question for yourself, once your Glowforge arrives, or with whatever laser you are building.)
The 20/100 was deep - almost halfway thru my quarter inch acrylic. I got the fuzz layer @Jules reported. Itās just the normal acrylic engrave residue (really makes a mess of lasers - ends up with a fine white dust everywhere inside and I usually clean the lasers after doing any significant acrylic work - itās worse with extruded). But after cleaning it from the piece the ridges are still there.
Looks like weāll have to wait for the engrave power tuning to come out from GF HQ before weāre going to get the truly frosted results.
Yes I quite understand. I might try and mock something up with my Shapeoku2, which has the same sort of axes, and a 1W laser diode. It should be able to engrave black acrylic and with X16 microstepping and 20 tooth GT2 pulleys get 2032 LPI and divisions there of. I have bottom of the range stepper drivers, mid range and high quality that I could connect to it. I also have a high precision XY table as a reference point. So I should be able to reproduce the same effect unless it is specific to CO2.
One questions for @dan is does the GF software try to engrave at the the LPI shown in the GFUI or does it always step a fixed number of microsteps? The GFUI LPI figures donāt seem to correspond to whole numbers of microsteps, so doing the maths in inches and then converting to steps would give aliasing. If it moves a fixed number of microsteps I get slightly different LPI numbers.
I believe the LPI were chosen because they corresponded to step increments, but it was a while ago that I last discussed it and my memory is fuzzy on the point.
They appear to roughly correspond to motor steps but not exactly because it looks like you use metric belts. If some high level part of the software actually used the LPI values displayed then it would give uneven steps due to rounding when converted to motor steps.