Taking an educated guess/looking at the pic, what I see is two fold... and possible three fold. Those marks are usually caused by belt joints. The two things that I believe are contributing to your issues are....
1. The build precision of the SG attachment (frankly, I have never see one of the SG attachments that could come even close to a stand alone Surface grinder machine in terms of flatness/precision.)
2. The hardness of the contact wheel you are using. In my experience, what you are seeing/experiencing, is worse with soft contact wheels versus hard ones. (Typical contact wheels that come with most grinders are 70 durometer. Sometimes you can special order a grinder with 90 durometer wheels, but I don't consider even 90 durometer to be hard enough for SG operations. (I personally use urethane on my stand alone surface grinder, with a hardness that is equal to 100 on the durometer scale, which is the highest/hardest the durometer scale goes)

Finally, the type of belt used for SG operations plays greatly into the finish..... I personally use either Norax or Trizact engineered abrasives for my Surface Grinder, and in doing so eliminate those "belt joint bumps".

Thank you!  I am indeed using a rubber wheel, so I should probably get something harder. I will start by swapping the belt for a Trizact ones to see how much of a difference it will make.

I sometimes see those marks with my SG attachment (Wuertz machine works) and I have found that moving fast makes them appear, while moving slowly eliminates them. I think Ed is right about it being the belt lap/joint that causes them. Set the motor as fast as it goes, and move the table slowly. See if that helps.

Oh, interesting. I was actually trying to do it as quickly as possible based on what the motion looks like in the YouTube videos. But maybe they just speed their videos up to make it look more interesting?  Thank you, I will definitely try that.

Ed,

Thanks for the tip about the Trizact belts for surface grinding. Makes sense.

-Evan

I second... (or third?) what these gentleman have pointed out on the belts. I found the best SGA performance with a nice new ceramic belt or a trizac. Normal aluminum oxide belts don't fit the bill for these SGAs due to the seam. 

When looking for the best "flatness" on my SGA I will advance the table towards the wheel only ~1-3 thousandths. Then make a very large amount of passes.... in the neighborhood of 20-30

Like mentioned above, these machines really aren't capable of much better than a +/- 3 thousandths variance over about 7'' in my experience. For making full tang knives and flattening Damascus billets, I think they fit the bill. 

I've toyed around with installing a solid metal wheel on mine instead of the contact wheel. This, I think, will make the machine less forgiving and make the surface grinding process more time consuming with smaller advances towards the wheel, however, I think the precision will be higher. 

My SGA:

[img] [/img]

Results with Test Piece and Trizac Belt:

[img] [/img]

Wow, you just blew my mind. I didn't realize that ceramic belts don't have a seam.  🙂

By the way, thank you for such an excellent video on how to make the attachment. It was a fun build and the resulting tool is extremely useful.

There are seams, but the "belt bounce" isn't nearly as bad as a J-Flex. Way smoother. I think doing the bulk of the flattening with ceramic, then hitting it with the trizac to dial it in on the SGA is a good order of operations. 

[img] [/img]

Cheers Luka!