All 4k @ 60 Hz input lag tests are done with 4:2:0, and all 1080p @ 60 Hz tests are done with 4:2:2, except when the test explicitly mentions that it's done in 4:4:4. Similarly, all HDR tests are done with 10 bit color except where explicitly mentioned. I've now added this information to the (?) tooltips of each test, which should help clear up the ambiguity.
Our (4k @ 60 Hz + HDR) test is done at 4k @ 60 Hz @ 4:2:0 @ 10 bit + HDR, but we also have a (4k @ 60Hz @ 4:4:4 + 8 bit HDR) that is meant for HDR PC use. Unfortunately 4k @ 60 Hz @ 4:4:4 @ 10 bit is beyond the bandwidth of HDMI 2.0, so no current TV will accept such a signal; 4k @ 50 Hz @ 4:4:4 @ 10 bit is technically possible over HDMI 2.0 (according to Wikipedia), but we currently have no equipment that can send this. Luckily 4k @ 30 Hz @ 4:4:4 @ 10 bit (or 12 bit) has been supported on every HDR TV a full-bandwidth port we've ever tried, so you can consider that a safe bet. Because of this you can consider our HDMI 2.0 Full Bandwidth test an equivalent test for what HDR signals you can send. Also I just added "Maximum refresh rate for 4k @ 4:4:4 @ 10 bit + HDR" to our internal Test Bench Suggestions list, we'll consider it during the next test bench update.