Looking at the datasheets, the ES56033 is nearly identical to the PT2399. Both use ΣΔ modulators for the A/D & D/A conversion. Both contain user-configureable low-pass filters. Both contain SRAM storage for the delay. Both contain a clock VCO. Both contain an internal bias supply. Both run on +5V. The pinouts are different. The SRAM depth is 32K in the ES56033, 44K in the PT2399. For the same clock freq, the PT2399's delay is 37% longer. Both datasheets leave a LOT to the imagination. The PT2399 provides a range for the clock freq and a table that maps resistor values to clock freq. The ES56033 tells us next to nothing about the VCO, other than two example circuits.
You mentioned that the clock freq changes when you touch the Time pot. If you swap the Time pot and the series resistor (18K in the example circuits), then you'll get some isolation between the pot and the more sensitive of the two FADJ pins.