Getting back to the original question
Hi Tom,
This isn't rocket science -
If you choose to use Peco turnouts - ALL Peco turnouts are designed to be fed from the single track going into the turnout. That's where you attach your feeds to the turnout - DCC or plain DC, it doesn't matter.
For Insulfrog turnouts - isolate the 2 rails leading back to the frog from the outgoing tracks. Power routing to beyond the frog is done under the frog area. Feed both tracks leaving beyond the turnout from the the other side of the isolation. Whether you use Peco's isolating track joiners or or use a Dremel to make the isolating cuts, it doesn't matter. Just isolate the 2 rails going back to the frog to prevent back feeding from the non-selected route if something bridges the insulation near the frog.
For Electrofrog turnouts - isolate all six rails going into the turnout, because the whole turnout is electrically live. These turnouts are not power-routing like the Insulfrog design. If all six rails are electrically isolated. Then your turnout is effectively an "electrical blob" matching the electrical "DCC polarity" of your selected route that won't short out as the train transits the turnout.
Yes, just isolating just the tracks going back to the frog should be sufficient, because the 2 outer rails through the turnout are electrically continuous. But for fault finding, if your turnout has its own feed and is electrically separate from all tracks, there is no chance an electrical fault elsewhere bleeding into the turnout, just as there is no chance of developing a track short caused by the track itself. That means if a loco hiccups when transiting the turnout, the fault is almost certainly with the loco rather than the track.
Some ancient history -
As I've said previously on here. (and been studiously ignored) The original Peco Insulfrog design from the 1970's or maybe even earlier had a massive solid plastic frog that had be "charged at speed" if you were to get over it without stalling. The Insulfrog design was later refined to minimise the amount of plastic frog and track insulation in the turnout some years back to get better slow-running performance.
Remember Peco is UK based. A lot of their UK prototypes are small 2 and 3 axle steam locomotives, with much shorter wheelbases than most US prototypes. This results in a smaller model that is a lot more electrically sensitive to non-powered track than longer locomotives with bogies and tenders to assist with electrical pickup when crossing a stretch of non-powered track. So the amount of insulated track in the Insulfrog turnout design has been kept to an absolute minimum.
Now if all wheels are properly tapered, it should be impossible to bridge across the frog area and cause a short..but.. Not all locomotives are set up that way. Once you go above 2 driving axles on a steam locomotive and certainly once you go more than 3 driving axles on a steam locomotive, then at least one of the axles will be fitted with so called "blind driving wheels" that do not have flanges and these blind drivers may be wider than the leading and trailing axles. On the prototype this is done to help get the long rigid wheelbase around curves. Model manufacturers have been using this trick for years to help our models get around our much tighter model curves. Also some model locomotive wheels are wider than "standard" to assist with tracking on our tighter model curves. By gapping the rails leading back to the frog, you defeat all of these problems that can cause a short in the frog area.
If you are running bogie diesels with all wheels picking up, the amount of plastic in the current Peco Insulfrog turnout design should not be an issue. You have the wheels on at least another 3 axles electrically feeding your locomotive motor when any one of the axles hits that now very small stretch of plastic isolation. The other 3 axles should be able to carry that load if your locomotive is correctly adjusted and cleaned. If you stall out - fix the locomotive, because the problem will be between your wheels and the motor and not in the turnout design.
If you are running small 2-axle industrial locomotives as I am, then when wheels on one axle hits the isolation areas, the other axle HAS to feed the locomotive motor or you stall out. Here's what I'm running successfully at low speeds
across an all Insulfrog exchange yard like this
The steam locomotive is built around a Bachmann Percy chassis. This is not a "Rolls Royce" chassis by any description, but if I can get Peco turnouts to work well with this type of rolling stock, then you should be able to as well.
As I said up front, It's not rocket science,