I don't like to leave my researching projects abandoned, so I just keep looking for an answer of what is happening with the histories.
It's very difficult to explain in just a few lines the strange things that I have seen. But I will try.

As I keep growing LookBack, probabilities of getting an error are higher and higher.

I've noticed that MT4 passes histories to Zorro in batches of 300 candles each time.
This is the log I get in MT4 with LookBack = 7000 and BarPeriod = 1, loading 2 assets:
Code:
2017.02.09 10:01:01.481	Zorro AUDUSD,M1: Zorro Disconnected!
2017.02.09 10:01:00.172	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.02 15:53 (6886 - 7186)
2017.02.09 10:00:59.828	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.02 20:53 (6585 - 6885)
2017.02.09 10:00:59.500	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.03 01:54 (6285 - 6585)
2017.02.09 10:00:59.188	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.03 06:53 (5986 - 6286)
2017.02.09 10:00:58.860	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.03 11:53 (5686 - 5986)
2017.02.09 10:00:58.563	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.03 16:55 (5387 - 5687)
2017.02.09 10:00:58.282	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.03 21:54 (5088 - 5388)
2017.02.09 10:00:58.000	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.06 03:04 (4788 - 5088)
2017.02.09 10:00:57.688	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.06 08:03 (4489 - 4789)
2017.02.09 10:00:57.406	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.06 13:02 (4190 - 4490)
2017.02.09 10:00:57.125	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.06 18:01 (3891 - 4191)
2017.02.09 10:00:56.844	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.06 23:01 (3591 - 3891)
2017.02.09 10:00:56.578	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.07 04:02 (3292 - 3592)
2017.02.09 10:00:56.281	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.07 09:01 (2993 - 3293)
2017.02.09 10:00:55.985	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.07 14:01 (2693 - 2993)
2017.02.09 10:00:55.672	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.07 19:00 (2394 - 2694)
2017.02.09 10:00:55.375	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.08 00:01 (2094 - 2394)
2017.02.09 10:00:55.110	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.08 05:04 (1795 - 2095)
2017.02.09 10:00:54.766	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.08 10:03 (1496 - 1796)
2017.02.09 10:00:54.485	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.08 15:02 (1197 - 1497)
2017.02.09 10:00:54.141	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.08 20:02 (897 - 1197)
2017.02.09 10:00:53.797	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.09 01:02 (598 - 898)
2017.02.09 10:00:53.438	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.09 06:01 (299 - 599)
2017.02.09 10:00:53.063	Zorro AUDUSD,M1: Send 300 USDCHF bars at 2017.02.09 11:00 (0 - 300)
2017.02.09 10:00:52.563	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.02 16:05 (6881 - 7181)
2017.02.09 10:00:52.203	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.02 21:04 (6582 - 6882)
2017.02.09 10:00:51.875	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.03 02:03 (6283 - 6583)
2017.02.09 10:00:51.532	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.03 07:02 (5984 - 6284)
2017.02.09 10:00:51.203	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.03 12:01 (5685 - 5985)
2017.02.09 10:00:50.891	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.03 17:04 (5385 - 5685)
2017.02.09 10:00:50.594	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.03 22:04 (5085 - 5385)
2017.02.09 10:00:50.281	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.06 03:11 (4786 - 5086)
2017.02.09 10:00:50.000	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.06 08:10 (4487 - 4787)
2017.02.09 10:00:49.719	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.06 13:09 (4188 - 4488)
2017.02.09 10:00:49.422	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.06 18:08 (3889 - 4189)
2017.02.09 10:00:49.125	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.06 23:08 (3589 - 3889)
2017.02.09 10:00:48.844	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.07 04:08 (3290 - 3590)
2017.02.09 10:00:48.547	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.07 09:07 (2991 - 3291)
2017.02.09 10:00:48.250	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.07 14:06 (2692 - 2992)
2017.02.09 10:00:47.969	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.07 19:05 (2393 - 2693)
2017.02.09 10:00:47.685	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.08 00:05 (2094 - 2394)
2017.02.09 10:00:47.364	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.08 05:04 (1795 - 2095)
2017.02.09 10:00:46.921	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.08 10:03 (1496 - 1796)
2017.02.09 10:00:46.530	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.08 15:02 (1197 - 1497)
2017.02.09 10:00:46.189	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.08 20:02 (897 - 1197)
2017.02.09 10:00:45.824	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.09 01:02 (598 - 898)
2017.02.09 10:00:45.449	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.09 06:01 (299 - 599)
2017.02.09 10:00:45.188	Zorro AUDUSD,M1: Send 300 EURUSD bars at 2017.02.09 11:00 (0 - 300)
2017.02.09 10:00:45.078	Zorro AUDUSD,M1: Zorro Connected!


The histories are there. I would not tell that the problem is in bad MT4 histories. The problem, I think, is in the DLL, in the "bridge" between MT4 and Zorro. I think that this bridge tries in some way to synchronize both timeseries, and because there are not exactly the same number of candles for the same period of time, the script disconnects in the MT4 end, and in the Zorro end it just says that it lacks X candles.

This is a problem.

Another one is that between different runs of the same script, sometimes it works, sometimes it doesn't, sometimes it says that it lacks X candles and other times it says it lacks Y candles.

I think that when LookBack is long enough and you are loading several assets, there should be a better check of histories correctly synchronized. This test could be done in the MT4 end (the DLL) or in the Zorro part. Personally, I would do it in the DLL... The sooner the problem is solved, the better.