Here's the thread. I never came to a successful conclusion. Even if I could get the times to match in both EKOS and the mount, the meridian flip would happen one hour early. I believe Somehow EKOS was compensating for DST even using only UTC time without an offset.
The only way I can successfully get flips to work on time, is if I let EKOS update the mount. I do have an MBox (regular) but GPS module on the 10Micron, which I use to create and build models. Basically I polar align the mount after GPS has updated it. Then load MountWizard4, and EKOS, which syncs time and location to the mount. Meridian flips work perfect this way.
If I set the mount to update EKOS, I have time mismatches between the hand control and EKOS, and I believe DST in EKOS still counts whether or not it's on or off on the mount. Then the flip doesn't occur correctly. Anyhow, feel free to look through the thread at what I tested.
indilib.org/forum/general/7928-why-is-th...y-mount.html?start=0
Another issue that crops up for me if I have EKOS update the mount with time and location. The second night out GOTO starts to get off by a few arc seconds. EKOS can barley nudge it back into position after plate solving. It saves, moves it a tiny fraction, then solves again, and continues to do this while it slowly nudges it over about 30 tiny moves until the object is centered again. The only way I've found to fix this, is to build a new model each night I use the mount. Annoying, but it only take about 20 minutes since it's automated. I never SYNC any plate solves. I only have EKOS update the time and location to the mount.