I've worked a lot with a table XRF machine and I'm telling you a handheld one is even more unreliable.
Even with my table XRF, I sometimes get false readings due to interference. For example, with presence of high amounts of Zn and W, it is impossible to determine whether there're Au or not. Especially raw ore usually contains like super little amount of PMs, 1% Au is like getting 1kg from 1ton. The so called high-grade Au ores we are dealing with in China are like 0.001% Au. So your ore grade is either making you millionaire or it's a false call. Your XRF result is showing at least 79.23% metal so it's (as mentioned above by others) almost metallic pieces. Smelt it and it will form alloy. However, according to the pictures, I think this is less like true.
My advice is to calibrate your XRF device first. To calibrate it, you need standards. To simplify, it's that you need to teach your device 'how does 99% Au look like' and 'how does 1% Au look like'. Then you make a super simple curve or linear coefficient. Then your device can estimate from 1% to 99% that how a 50% Au would show. Doing such calibration would give you semi-quantitative result. I expect Au grades usually lower than 0.05% and most likely lower than 0.005%. If it's such case, even 1% and 99% standards won't be very helpful. Because Au doesn't come alone in raw ores, it comes with other base metals. 1% Zn and 1% W can completely 'cover' 0.1% Au and you won't see anything to indicate Au from the spectrum.
The more standard way to use XRF device is to determine a sample's main contents. Like, when you find Sb, it's most likely with Au. When you find As, it's most likely with Au. Also, you get to know what to remove, so that to 'concentrate' the PMs. Or, you can use ICP, etch what you want and read the solution.
I feel like I'm leaving the subject. Anyway, for what you have there, it's impossible to do a proper and reliable assay. You better send it to qualified lab or get some other things to tune your XRF device.