I created a complex EA with more than 6 variables on entry/ exit , on time frame M1.
Loaded the EA into MT4 and MT5 respectively and properly.
The EA's performed contrary to FSB results , very poorly. ( Not expected to work anyway due to the fact that the back-test  was null or not successful)
I used 4 different demo accounts : IG, FxChoice and Alpari for MT4 and Alpari for MT5,  it did  run in extremely slowly on all 4 demos. (The CPU  and memory of the computer remained under 60%)

I observed:

1.- The code is approx. 8000 lines long.
2.- It takes for ever to do a back test of a few days (More than an hour in a very fast machine 4Gz ( 8 cores), & 32Gb of Mem)
3.- Even thought the back test was initiated in MT5 , it was  stopped several times after a longer testing period was entered making the back test even more cumbersome and lengthy in time (hours) - Usually MT5 uses all the 8 CORES during  this  process but it seam for ever the waiting time.
4.- In MT4 the back test end up with results not in accordance to FSB modeling.
5.- In FSB the EA works like a champion.
6.- After running Montecarlo and Comparator , still the EA looks like a charm but ot after back*tested on MT4 .
7.- The code compiles 100% good in MT4 , MT5.

I am submitting the FSB and the EA for MT for further analysis of the case and find a resolution.
Most probably I am doing something incorrect, but I do not know at this time .
Possibly too many variables are been used!

Txs.