Hi Tim,
Seems to me like you've already received a bunch of good information. Let me add my two cents worth ...
Taking your questions in order (but only so that I can digress!)
1) First, i have to assume that you are using the new triton EQ that has the balancer built-in. True or no? As to how accurately it measures the charge put back in, I can't say authoritatively: Electrifly do not even publish a specification for this. On the other hand, I'd be surprised if it wasn't accurate to within just 2 or 3%, just based on the needed voltage measurement accuracy and the moderate cost of 1% tolerance components.
2) The LiPO charge cycle has two distinct stages: in the first, the charger "tries to" supply 21.0 Volts (5S) but it cannot because the adjustable current limit gets in the way. This is the Charging Current setting that you chose for nominal 1C charging rate. Eventually, the pack achieves 21V but it is still charging. The current now begins to "tail-off" from its original limit and the remaining 10% of charge will take maybe one-third of the total charging time. The charge is actually declared complete when the current tail drops to maybe 1/50th of the initial current.
2b) If your charger balances the pack, then balancing ideally takes place during this second phase of the charge cycle.
2c) If the pack is substantially unbalanced, then the first cell will achieve 4.20 (really 4.22 for some chargers) Volts before the entire pack has reached 21.0V. If
and when this happens, the charger will (should / must?) "throttle back" to a charge rate that is no greater than the "thieving" or bypass current. In the Triton specifications sheet it is called the "node current". The thieving current is
steered around the fully/overcharged cell so that it does not charge any more but the other cells still get charged ...
BUT AT ONLY 300 mA! If you were charging your 3000 mA-h pack at 3 Amps before, now you are charging 1/10th as fast. This substantially lengthens the time needed to finish phase 1 of the charge, if the cells are unbalanced more than 0.01 or 0.02 Volts.
2d) ALL balancing chargers on the market do this, and must do so, unless they are built using a dramatically more expensive architecture.
2e) 2C, 3C and 4C charging will lose their time advantage if the cells become unbalanced, and the fastest way to unbalance them is to ... wait for it! ... charge them really fast. The other ways to unbalance them are to run them down past 80% capacity and to get them too hot.
3) Yes, 90% discharge is too deep. You will probably need to use bigger batteries ... but let's not get too far ahead of ourselves.
4) Changing the pitch and running RPM may (or may not!) do what you want. In general, motor efficiency suffers when you drop the running RPM but prop efficiency improves a little. For minimum battery consumption, you want to drop pitch and raise RPM until you barely have enough voltage left at 75% discharge to maintain lap time. (If you have a Castle ICE you'd see the %DRIVE approach 95% just before the clover)
Here we need a whole lot more information from you: plane weight, line length, how draggy is it, motor, ESC, battery and desired lap time.
For starters, get the plane slowed down to the desired lap time. This has a dramatic effect. Then if the results are encouraging, sneak up on the flight time. Also try to trim the wasted time off the end of the flight, and lengthen your walk-to-the-handle delay so that you waste as little capacity as possible. The downside of dropping the pitch as far as possible is that you lose punch in the hourglass/overhead 8 / and clover when you don't have excess voltage available at the end of the flight.
5) Yes! Like I said, slow the lap time first then sneak up on the flight time.
6) I'll say it again, we need every bit of info you can think of. An explanation of why the sky is blue would be helpful too!
I think you see from the enthusiasm with which answers have accumulated, that we all want you to succeed, so let's git to it!
Dean