Add Gaussian quadrature to some of the examples with monte-carlo expectations

[jlperla]

Probably stick with gaussian quadrature, maybe in the growth model and only with normal distributions? Just go nodes, weights = gausshermite(n)

Can also show the adaptation for a normal that isn't

function gaussian_nodes_weights(mu, sigma, n)
    gh = gausshermite(n)
    nodes = gh[1] .* (sqrt(2) * (sigma)) .+ mu
    weights = gh[2] ./ sqrt(pi)
    return nodes, weights
end

or something like that.