Started with a 4X8 table that we had as a work surface at the shop.
Figured that a 8 X 11 frame worked slightly better than an 8X12 based on the excellent frame checker:
This meant that the angle of the arms needed to be 53.1 degrees.
So, I cut out a template on my small shapeoko and made a cutout in each of the corners using a router:
Installed some machine inserts and used some aluminum 8020 as arms, but the flex was way to great since the arms were 30" overhang.
So, made some wood reinforcement that the 8020 could sit in to stiffen it up, and shimmed where necessary to get a very tight fit against the table.
Results are not totally amazing, managed to get a .47 using the default 9X9 2000X1000 grid.
Couple of questions.
First, and most obvious, is that good enough to cut a full 4X8 sheet, and any suggestions on improving? I used that standard tensions, didn’t really change any settings except for the initial size measurement. I know I can just modify the acceptable calibration floor to get it to pass. 
Second, what are the benefits of using a larger grid vs a grid with more points in calibration? ie, is it better to maximize the size of grid, or increase the number of points. I have been scouring the forums for an answer that explains the relationship, but haven’t found one. If someone has a link to share that explains better, would love to see it.
Third, the frame is now quite stout in the x,y plane, but there is a bit of flex in the z direction due to the long moment arm. Is this an issue? My main goal is to be able to install/uninstall by just loosening the 2 screws holding the 8020 and sliding it out, as the arms really get in the way when not in use. I don’t see an obvious way to hold down the z without attaching it to the table with an angle bracket.
Fourth, I saw there was a frame flex check available with the 7.4.1 firmware. I didn’t really want to reinstall firmware, but I didn’t see this was incorporated into later versions. Am I missing it? Was wanting to see how stiff my frame was in relation to others.
Fifth, I saw that a .5 calibration means the cuts will be within 2 mm accuracy. I think I saw that .4 calibration results in a 3 mm accuracy? Is there a published table somewhere that shows accuracy as this number changes?
Finally, I would like to be able to move this from the shop at work to my garage at home where I plan to install concrete anchors, is there a way to save calibrations? Would it be as easy as having two maslow.yaml files and just renaming them depending on which one I want to use?
Other than that, the build directions were really excellent, my only questions that weren’t addressed in them were what to do with the crazy long belts. It wasn’t obvious that all that extra belt would wind up correctly, and I almost trimmed them, but saw a post recommending against doing that in case you wanted a larger frame. Once I hit the retract all I saw that it seemed to spool up pretty good.
Thanks for making such a great product. Not entirely sure what we will do with it, but it is really cool.
