Right, let's do this systematically:
1. That second inversion wasn't an Immelmann Loop. An Immelmann loop starts with the half loop, then dives out to the right or left, just like SheiKra (
http://www.rcdb.com/2662.htm?p=12151). Your inversion was a Dive Loop, where the train climbing upwards then rolls into the half loop and out the other side, like on Mantis (
http://www.rcdb.com/7.htm?p=34).
2. Try to use three segments for the lift hills. The first one goes from the horizonal to the desired angle for the lift. The second segment is straight (select the segment and press 'I'), then the final one levels off the top of the lift. Makes for a much more realistic lift hill. To use Mantis again (I still have the tab open) you can see it's lift hill follows this pattern:
http://www.rcdb.com/7.htm?p=2621
3. Generally there were banking problems all the way round the ride. The worst one was the Dive Loop, try using the 'Rel-Roll' button next time. This basically means you bank the segment of track in relation to the previous node, instead of some global position. This means that you can avoid the banking problems with vertical track.
4. Just before the MCBR there was a jerk in the track, I suspect this is because you tried to straighten the MCBR without realigning it. Basically just one to look out for.
5. The supports on the loop were well out. Again, Mantis will serve as a useful example, so check out this picture:
http://www.rcdb.com/7.htm?p=29 . Basically, if you make your supports like this, then it's a lot more realistic.
6. The bunny hills were awful. Sorry, but they were. They were pumpy, jerky, the pull outs were too sharp and there was even a wiggle in the track on one of them.
7. The flat s-bend before the final brakes was a bit stupid too! Why didn't you just extend the brakes? The laterals would have been awful.
8. I can't see the forces on the video, but I'm willing to bet they're all over the place.
So yeah, I've not got much positive to say I'm afraid, but keep at it. You'll soon get the hang of NL, it just takes practice. And feedback.
