I have a good excuse for not replying earlier
.
The torque converter used in the Powerstroke is designed simply to "save the transmission". In other words, it has a high stall speed and a very low torque multiplication factor. This allows the engine RPM to rise very high in relation to the actual input shaft speed. If you factor in the transmission gear ratios (which are absolutely stupid and useless in a 4-speed automatic) of 1st (2.71:1), 2nd (1.54:1), and 3rd (1:1), then it becomes apparent that the act of going from 1st locked to 2nd unlocked will actually cause an INCREASE of engine RPM during a shift....and a shift that causes a near absolute loss of drive wheel torque. The same would happen (and even moreso) during a 2nd locked to 3rd unlocked transition. The limited and closely spaced gear ratios in the E4OD and 4R100 make this a very undesirable mode of operation and once the torque converter is locked, it makes sense to keep it there.
The biggest problem I have noticed, especially when using the stock PMT2 calibration that my truck came with, is that the torque converter unlocks for a full 2.5-3 seconds during a 4-3 downshift while under load either forced or after a TCS (OD off button) push. It's ridiculous how much rear wheel torque is lost during this short time....and it's extremely annoying. The problem is compounded by an instant jump to 3200 RPM at 73 MPH....if you look at a dyno graph for a 7.3L, you'll see that 3200 RPM is useless. Good thing there are ways around it with different PCM codes and programming techniques.
I would be VERY hard to get the shift timing right while attempting to tune what you are describing. This is mostly due to the variable shift points that are encountered depending on acceleration rate and different accelerator pedal positions.
On the white smoke issue, my thoughts are that of Bill's. I would also assume that it is due to the slightly higher commanded injection pressure combined with the added cold start of injection that cleared up the white smoke. I will also point out that white smoke on startup can also be caused by low compression or an inoperative glow plug or two. But since it didn't smoke when the ICP sensor was unplugged, I don't think I'd blame any mechanical issues.
The PCM will default to 725 (calculated by Injection Pressure Regulator duty cycle) at idle when the ICP sensor is unplugged. This is a bit higher than the 485 or so PSI that is commanded when the ICP sensor is functional.