The original Centipede was built (using the chassis of the Essl locomotive) for SAL, to do the work of a good 2-6-6-4 in high-speed freight service.
The basic premise of the Centipede was to get GG1/T1 train-handling power in a locomotive one union crew would operate. The construction had the electrical cables neatly packaged under the floor, with no drains, a perfect trap for the oil and coolant that leaked out of the many hose connections.
Pusher service was exceptionally woeful for reasons I think could have been predicted. On sharp curves the traction-motor blower ducts did not line up between the body and the underframes... just at the time hourly-rate cooling was most desired. There were EIGHTY brakeshoes on one unit, which in pusher service was often operated light downhill entirely on the independent brake -- these were No Fun To Change (and we have at least one jeremiad about just how bad this was!)
Three BP-20s represented a better way to make the 6000hp by 1948...
The centipedes were two separate locomotives. Most of them used a drawbar to attach the two units together. I'm not sure which road used a coupler on the rear end. I know the PRR used the drawbar and considered both units as a single locomotive. 6000 HP when built but later knocked down to 5000 HP. They were really hoping that they would work out in passenger service but problems relegated them to freight service after a gearing change. Eventually they wound up in pusher service and snappers. Too bad they didn't work out. Very unique locomotives.
Pete.
They only had 'cabs on both ends' because they were two single-cab units coupled back-to-back to make a "6000hp locomotive".
Kiefer noted in the 1947 survey of motive power that a 'single Centipede' was the shortest way to provide 3000 diesel horsepower at that time -- everyone else needing multiple units taking up more track length. I believe two back-to-back would fit a 120' turntable, if you had some need to turn them, but the whole point of having a cab at both ends was to simplify bidirectional operation.
The double-cab engines you're thinking of were CNJ's 2000hp commuter power, and they rode on conventional A-1-A trucks. They were worked the same way commuter tank engines were, on comparable services. (You could probably turn one on a table capable of turning one of the camelback 4-6-0s with tender... but why bother?)
'
Since these things had cabs at both ends on the PRR I assume there was never a reason to turn them. If they were turned I also assume no turntable existed they could fit on and would have to be turned on a wye somewhere. Any thoughts?