Vector analysis in curvilinear coordinates

It is known that vector calculus in 3D takes quite a simple form when one uses orthonormal curvilinear coordinates (check out, for example, here, in section "Differentiation"). Is there any possibility to easily use this approach in Mathematica currently (giving just metric tensor $g$ and using some kind of "Laplacian($f$, $g$)" and "Gradient($f$, $g$)", for instance)?

If you are satisfied with the metrics defined in CoordinateChartData[]:

CoordinateChartData[] //Short

{{{Bipolar,{\[FormalA]}},Euclidean,2},{{BipolarCylindrical,{\[FormalA]}},Euclidean,3},<<62>>,{Stereographic,{Sphere,{\[FormalCapitalR]}},10}}

then you can do things like:

Grad[f[r, θ, z], {r, θ, z}, "Cylindrical"] //TeXForm

$\left\{f^{(1,0,0)}(r,\theta ,z),\frac{f^{(0,1,0)}(r,\theta ,z)}{r},f^{(0,0,1)}(r,\theta ,z)\right\}$

and:

Laplacian[g[r, θ, ϕ], {r, θ, ϕ}, "Spherical"] //TeXForm

$\frac{\frac{g^{(0,2,0)}(r,\theta ,\phi )}{r}+g^{(1,0,0)}(r,\theta ,\phi )}{r}+g^{(2,0,0)}(r,\theta ,\phi )+\frac{\csc (\theta ) \left(\sin (\theta ) g^{(1,0,0)}(r,\theta ,\phi )+\frac{\cos (\theta ) g^{(0,1,0)}(r,\theta ,\phi )}{r}+\frac{\csc (\theta ) g^{(0,0,2)}(r,\theta ,\phi )}{r}\right)}{r}$