Do not define `Base.in` instead of `Base.issubset` (taking set theory seriously)

The definition at

https://github.com/JuliaGeometry/GeometryBasics.jl/blob/234699e015ccb23bdaab26539dfc8fdfa9ff6add/src/primitives/rectangles.jl#L469-L481

is not in the spirit of Base.in, and instead that procedure should be a method for Base.issubset (⊆).

see for example:

https://github.com/invenia/Intervals.jl/blob/168565ef0a40d022bfe4030d0178cb1b83fff6cb/src/interval.jl#L370-L375

Base.in(a, b::AbstractInterval) = !(a ≫ b || a ≪ b)

function Base.in(a::AbstractInterval, b::AbstractInterval)
    # Intervals should be compared with set operations
    throw(ArgumentError("Intervals can not be compared with `in`. Use `issubset` instead."))
end

These operations are related, even graphically (∈ ⊂), but should not be mixed up.

Start at src/primitives/rectangles.jl lines 469-481 and compare the current Base.in definition with the referenced Intervals.jl implementation. Check Julia's Base.issubset and Base.in conventions, then update the rectangle operation so set containment uses issubset rather than in. Done means the API follows the issue's set-theoretic distinction and the existing behavior is covered by the relevant project tests.