Fix for unhandled TKOFF_THR_MAX

[Georacer]

This patches a bug that was introduced via #27174.

It fixes an unhandled case of TKOFF_THR_MAX=0, where THR_MAX would not be sourced as a value for the minimum throttle limit during takeoff.
The relevant conditions are:

• TKOFF_OPTIONS=0 (default)
• TKOFF_THR_MAX=0 (default)

This would result in incorrect backing-off of the minimum throttle during the climb, which @Hwurzburg spotted:

Now it behaves as it should:

Log: https://www.dropbox.com/scl/fi/akdqr9o5niv3qleexh3yd/122221_TakeoffTakeoff4.bin?rlkey=o9uykk6cm7785oxctkj17txxq&dl=0

---

Side-note:

I notice that there was significant pitch and throttle oscillation in Henry's log.
I don't think this is related to #27174, as the oscillation persists even after TAKEOFF mode has finished climbing, where TECS operates in a regime which I haven't edited.
I believe that a sudden pitch-down and throttle-cut was triggered (because of this bug) during the climb and the default tuning and model physics would yield bad control behaviour.

[Hwurzburg]

This makes it safe from my sitl testing...
in another PR we need to get the model fixed so as not to go unstable
and , more importantly, get the takeoff leveling working again to prevent the overshoots in altitude on auto takeoffs

[peterbarker]

Pretty sure this is breaking several Plane autotests.

[Georacer]

@peterbarker I believe the latest commit fixed it.
It was a case of the minimum throttle during a transition being set to TAKEOFF_THR_MAX. This is because in servos.cpp the transition throttle limits were historically being treated the same as the takeoff throttle limits. And this resulted in a back-transition getting a minimum throttle allowance of TAKEOFF_THR_MAX.

For now I have reverted to the behaviour before #27174.
@tridge I'll have to find another way to introduce the minimum throttle during a transition (and find a way to tell between a forward and backward transition).

The failing autotests passed locally, but let's see what the CI server says about it.

[tridge]

i think we can make min takeoff throttle work for fwd transition