failfactor in IntegratorOptions not used and causing type problems

[kylebeggs]

The failfactor field doesn't seem to be used and causes issues when trying to use solution types other than Float64. Either we can remove it, or create a type parameter of its own so it does not need to match.

Thunderbolt.jl/src/solver/time/integrator/type.jl

Line 13 in 016a8d2

failfactor::tType = 4.0

[termi-official]

We can comment it out for now (and same goes for other unused stuff).

For some context here, I am working on how to adjust the internal dispatches in OrdinaryDiffEqCore to make external integrators compatible with the solvers, so I just mirrored some parts, but ended up not using others right now, as I do not fulfill the complete interface yet. See SciML/SciMLBase.jl#896 for a starter.