I really like conforming/broadcasting, especially when it is explicit and when the scalar version still makes sense.

I wonder if Lil allows conforming 3 or more lists together. If it does, what is the rule for which lists to reshape?

Here’s how explicit broadcasting can work in Julia

# A naive, scalar linear interpolation function
julia> lerp(x, y, z) = x + z*(y - x)
lerp (generic function with 1 method)

julia> x = [10 20 30]
1×3 Matrix{Int64}:
 10  20  30

julia> y = [40 10 20]
1×3 Matrix{Int64}:
 40  10  20

# A range literal, converted to a column vector so that you can see it
# is a different shape than x and y. Need row and column vectors
# because that's how matrix math works.
julia> z = [0:.2:1;]
6-element Vector{Float64}:
 0.0
 0.2
 0.4
 0.6
 0.8
 1.0

# Doesn't work because z*(y-x) is a 6x3 matrix and you can't add a 1x3 vector
# to that because addition is undefined for matrixes with different shapes by
# the conventional rules of linear algebra.
julia> lerp(x, y, z)
ERROR: DimensionMismatch: dimensions must match: a has dims (Base.OneTo(1), Base.OneTo(3)), b has dims (Base.OneTo(6), Base.OneTo(3)), mismatch at 1
Stacktrace:
 [1] promote_shape
   @ ./indices.jl:178 [inlined]
 [2] promote_shape(a::Matrix{Int64}, b::Matrix{Float64})
   @ Base ./indices.jl:169
 [3] +(A::Matrix{Int64}, Bs::Matrix{Float64})
   @ Base ./arraymath.jl:14
 [4] lerp(x::Matrix{Int64}, y::Matrix{Int64}, z::Vector{Float64})
   @ Main ./REPL[34]:1
 [5] top-level scope
   @ REPL[35]:1

# But you can get the right result if you explicitly broadcast by adding a dot
# before the parenthesis.
julia> lerp.(x, y, z)
6×3 Matrix{Float64}:
 10.0  20.0  30.0
 16.0  18.0  28.0
 22.0  16.0  26.0
 28.0  14.0  24.0
 34.0  12.0  22.0
 40.0  10.0  20.0

# Or the function author could have written a function that works for both the
# scalar and non-scalar cases by broadcasting the addition and subtraction:
julia> lerp(x, y, z) = x .+ z*(y .- x)
lerp (generic function with 1 method)