The only P-38 that I have had experiance with is a Royal that was grossly overpowered (the worstcase senario for single engine handeling). This airplane was set up for warbird pylon racing and I only got perhaps twenty flights before it was lost not to an engine out but the result of a rather heavy wing loading and my over exuberance pulling G's at the #1 pylon with the resultantant instant snap roll into the ground. I only had one engine out on an occasion that happened at speed so was not a control issue.
This ship had no gyros and twin 'Y' corded rudder servos with equal throws. As far as the full scale having rudder differential, perhaps but I doubt it. My thoughts on that is particularly with an aircraft with minimal vertical stabilizer and rudder area you certainly do not want one of those rudders in an asymetric thrust situation to be capable of only partial throw even if it is the one operating in the relative low energy air behind the dead engine.
As far as engine out recognition and control goes and this applies to any multi configuration to various degrees except 'inline thrust' there are four possible 'crutchs' or pilot aids if you will that serve the purpose of buying time for the pilot to recognize an engine out situation and to respond They are:
1. Closely spaced engines. Not always possible with scale.
2. Equal out thrust on both sides. Very effective method that is not hard to hide for scale purposes but in many cases this is prototypical for the full scale.
3. Rudder gyro. Absolutely a very effective device to buy time.
4. Horizontal stabilizer tip plates. A little known device while not as effective as the first three above for buying time but none the less does help and has been well tested on my Duelist. These are simply small plates on the tips of the stabilizer with a built in deflection built in i.e. right plate applies a right rudder imput and the left plate has a built in left rudder imput. Both plates are of equal and opposing deflection. Zero yaw in normal conditions. (yes there is a small total drag penilty to pay. When an engine is lost say the right one, there will be a loss of the high energy air flow over the the stab tip plate on that side which is behind the dead engine. This results in that plates inability to balence the left one and the result is always a yaw imput into that good left engine with no thought or reaction imput which is exactly what you want. Think what you may, this does buy that precious time and is well tested on my duelist and no it was not my idea.
Now having said all that I have never used gyros on a multi and simply because I was being cheap however I consider Twinmans rather rare research in the matter for our application solid gold.
One last thought after you are already in that grave yard spiral is not the time to start thinking about what to do and precisely why useing one or all of the above time 'crutchs' above are so very important.
Just returned this morning from a delightful flight with the Quad Kaydet.
John
Royal P-38, deceased.
Smallish UC-78 scratcher w/.25's.
World F-82, deceased from structural failure and wing folding in pylon racing.
Two bashed twins from trainers w/.20's and .25's.
A bashed quad from a trainer .25's.
My design eighty inch Stinson SM-6000 triple w/.25's (unfinished through cover).
Vintage Douglas B-26 Ukie, unthrottled Fugi .29's.
Duelist, Rossi .45's earlier now TT.46's.