Having gotten two answers from two USB meters, I figured it was time to get primal:
That’s a pair of USB breakout connectors and lengths of nice silicone wire (24 AWG power & 28 AWG data), with just enough slack for a Tek A6302 current probe:
So I can see the actual current waveform of a Glass Tile box running from a bench power supply:
The top trace is the firmware heartbeat from the Arduino Nano, the middle trace is the SK6812 LED data stream, and the bottom trace is the USB current at 50 mA/div. The current steps downward by about 10 mA (just after the data burst) when one of the tiles changes color and and LED shuts off.
The current probe reveals some mysteries, such as this waveform from a dirt-cheap USB charger:
I wonder why it’s ramming 100 mA current spikes into the circuit, too. At least now I can see what’s going on.
The Smell of Molten Projects in the Morning 
Obviously it’s time to bust that dirt cheap charger open and see what’s in there. My guess is a resistor and seven silicon diodes in series, because Zeners cost money, people!
Good thing I got ten of ’em, isn’t it? [grin]
The way the wires cross over each other in that first picture really had me scratching my head for a minute.
Yeah, that was the third attempt. The first two pix were even worse!