High DPC Latency will affect *any* audio interface (doesn't matter if it's PCIe, USB, Firewire, or Thunderbolt).
To effectively work with audio, the machine needs low/consistent DPC Latency.
 

So I made sure the 2i4 was the only USB plugged into a section. MUCH better, but still have the crackles. I used Windows Performance Recorder to make a trace file in WPA. Which DPC status do I look in?
DPC/ISR by Function?
DPC by Function?
ISR by Function?
DPC Timeline Module?
 
 
Can I make a copy of this data to show you guys?
 

I have an ASUS Z170 motherboard and do not have any DPC Latency problems.  

We used Asus Z170 motherboards as well... but they made numerous Z170 motherboards.
It's possible that particular motherboard has a problem.
It's somewhat rare, but I've seen it before.
Years ago with an Asus or Gigabyte motherboard... and more recently with EVGA.
 
If the OP has ruled out literally every other piece of hardware, the motherboard is the only thing left.

I ran another Latency test, this time at 512 Buffer Size, cleaner sound still with crackle, though not as much and found 3 Drivers with high execution.
 
WDF1000.sys
- ISR Count: 20014
- DPC Count: 20012
- Highest Execution: 0.308
 
HDAUDBUS.sys 
- ISR Count:16054
- DPC Count: 1428
- Highest Execution: 0.144
 
DXGKRNL.sys
- ISR Count: 14626
- DPC Count: 8000
- highest Execution: 0.108
 
 
Before we seal the coffin, any thoughts here?

FWIW, The ASIO buffer size of your audio interface has no bearing on DPC Latency.
 
If your DPC Latency is a littler on the higher size (say 500uSec), running larger ASIO buffer sizes can mitigate that... at the expense of higher audio latency.
 
To effectively work at low audio latency, you have to have low/consistent DPC Latency.
 
If you've gone thru all the other variables, I'd swap out the motherboard.
 

One thing that might be worth trying and is cheaper than a new motherboard is a PCIe USB card that has only the audio interface connected. I found that solved a dropout problem with a Cakewalk UA-101 interface a build or so ago. For whatever reason the m/b USB was fine with loads of MIDI stuff hooked up but I couldn’t get latency down below about 512 samples at 44.1KHz even if the only thing connected was the interface. Connecting the UA-101 to a dedicated USB board let me drop to 64 samples without problems.

FWIW, If the machine is suffering from particularly High DPC Latency, it doesn't matter what type of audio interface (or how it's connected).  Audio glitches will happen because the CPU is monopolized... and can't get the ASIO buffer filled in time (thus the pops/ticks and even drop-outs).
 
If a USB audio interface is sharing a root-hub with another device, that can affect performance.
If a USB audio interface is connected to a 3rd-party USB-3 controller (not an Intel USB-3 controller), that can cause a problem (especially with older USB-2 audio interfaces).  Many older USB MIDI interfaces have trouble with 3rd-party USB-3 controllers.
 
 
 

wdf might be part of a driver stack for some piece of hardware in your system. "The web" says many specialized mouses/trackpads use it. For starters, you could go to your device manager and go through each device, starting with mouse and display adapter & anything associated with firmware or BIOS or motherboard manufacturer, do properties, then the Driver tab, then the Driver Details button and look for the Wdf01000.sys entry in the list. 
 
I didn't find it in either computer I looked at, so maybe it won't show in those lists, or maybe not every computer uses it. Edit: wdf01000.sys appears in the driver list from LatencyMon, just not in any hardware driver details I looked at. OTOH it only had 6 ISRs in 30 seconds.
 
Any antivirus involved? Windows update working (Background Intelligent Transfer Service)?

Jim Roseberry
FWIW, If the machine is suffering from particularly High DPC Latency, it doesn't matter what type of audio interface (or how it's connected).  Audio glitches will happen because the CPU is monopolized... and can't get the ASIO buffer filled in time (thus the pops/ticks and even drop-outs).

Agreed. Except that build didn’t have high dpc latency, at least it didn’t so long as wi-fi was disabled - and as often seems the case the ethernet chip on the board worked fine dpc wise. Driver problem? Problem with the m/b USB circuit or drivers? I don’t know, but putting the interface on a seperate USB card solved the problem. One of life’s mysteries.....