What is wrong with the cell wall idea. Yes it is a lipid double layer, but it seems to separate the inside and outside quite well (you need all those channels to get stuff in)

As a complete layperson, I suspect osmosis may be a good counterexample. Channels react selectively to various compounds, but water can probably pass through the membrane more easily.

Wikipedia (heh) seems to support this idea, but my gosh is this all complex.

> Compared to ions, water molecules actually have a relatively large permeability through the bilayer, as evidenced by osmotic swelling. [...] Small uncharged apolar molecules diffuse through lipid bilayers many orders of magnitude faster than ions or water. This applies both to fats and organic solvents like chloroform and ether. Regardless of their polar character larger molecules diffuse more slowly across lipid bilayers than small molecules.

https://en.wikipedia.org/wiki/Lipid_bilayer

But then any model has its limitations, doesn't it. I think hydrogen can go through metal (and probably walls), so you could say walls are not a good model for walls.

Yea,those channels. They're complicated and x ray crystallography only broadly supports a thesis they are little holes. There's all kinds of weirdness going on, opening and closing by twisting under voltage and other stimuli. If you think it's a little pipe or even like a stomata (which is a macrocellular structure in plants which have cell walls in real sense) you've been led down the metaphor too far. Or, so I am told. I'm not a cell biologist.

And that biology in general is extremely random. School teaches it as a bunch of little machines.

Evolution should have been taught as applied statistics and stochastic effects. Using language of implied determinism by genes was a suicide move to the religious. It's applied probability folks, "more successful" is on average. Selection pressure is on average.