After my last test I conclude that to make the robot go faster, I need to use a higher current rated battery to pump more current into the coils, or use a lighter one to.

But what if I reduce that weight by cutting off two legs? And that's what I tested. By doing this the robot was no longer omni-directional, but this change immediately showed that the concept was working better. The magic happened after shifting the turning axes of the robot. Its not 100% balanced, from the video you can see that one of the legs is slightly shifting and not move straight. But it's now probably faster than a turtle :-)

I'm not going to abandon the 4-legged PCB Robot design just yet. I still looking and get quotes from different manufactures on different lipo batteries. Once I receive the battery I will test the robot again and maybe even add a similar pivot to see what happens. In the meantime, I I can also start playing around the the modules Bluetooth stack!