True and in instrument procedures, the weather is critical. If the ceiling is 400 with 1/2 vis, and you find it is 1,000 ft and 1 mile, that ruins the entire experience. 

 

I think it can be realistic... sometimes the weather is changing rapidly anyway.  Or imagine the fun of going missed when your Wx is lower than planned hehe.

Highly agree about the weather and horizontal layer modelling. I was programming my CDU to capture the ILS at Seattle, and there were clouds all the way to 500 ft. From a Cat III all the way down to Cat I. Extremely jarring. But this is also why I say X-Plane's default cloud rendering system looks extremely good for stratus and cirrus, but not cumulus-type clouds.