`dispersion_parameter(::Geometric)` incorrectly returns true

[palday]

GLM.jl/src/glmtools.jl

Lines 484 to 500 in 0e145bd

	"""
	GLM.dispersion_parameter(D)
	Does distribution `D` have a separate dispersion parameter, ϕ?
	Returns `false` for the `Bernoulli`, `Binomial` and `Poisson` distributions, `true` otherwise.
	# Examples
	```jldoctest; setup = :(using GLM)
	julia> show(GLM.dispersion_parameter(Normal()))
	true
	julia> show(GLM.dispersion_parameter(Bernoulli()))
	false
	```
	"""
	dispersion_parameter(D) = true
	dispersion_parameter(::Union{Bernoulli, Binomial, Poisson}) = false

The only parameter of the geometric distribution is the success probability.

[andreasnoack]

I'm wondering if we dare define this as length(params(D)) > 1? Any obvious cases where it would be wrong? Alternatively, we'd probably have to define the list explicitly and error out for unknown distributions.

[nalimilan]

I find it better to only define methods for known distributions and let users define their methods for custom distributions. It's just too risky to make assumptions. See #624.