If I could make a suggestion -- use lye concentration or water:lye ratio and avoid using "water as % of oils". As you change the fats in your recipe, controlling your water with the "water as % of oils" setting will not give you a consistent lye concentration. This variability will give you less consistent soaping results.
So, back to your question -- I agree with IL; I use a 25% lye concentration (3:1 water:lye ratio) too. It seems to be a sweet spot that works well for most people.
I inadvertently did a LS recipe at 33% lye concentration and found the soap paste to be very thick and difficult to stir and such. But it worked as far as saponifying well -- just hard to deal with.
We've also talked about LS recipes that use more water than the usual 3:1 ratio. I think the consensus of those who have tried more water is that it can be tough to get the soap batter to form a stable emulsion. If I remember correctly, folks started to run into that problem with a 20% lye concentration (4:1 water:lye ratio). I can confirm from personal experience that a 15% lye concentration is difficult to work with.
I usually use a glycerin and water blend as my water phase rather than all glycerin. I dissolve my KOH in the same weight of water, and then add any combination of glycerin and water to make up the rest of the liquid I want in the recipe.
Example:
My recipe calls for 25% lye concentration. That translates to a water:lye ratio of 3 parts water (or water and glycerin) and 1 part lye.
It also calls for 100 g KOH (to pick a nice easy number).
I dissolve the KOH in 100 g of distilled water. There's one part of the water phase that I need.
To make up the other two parts needed, I would then use one of these: (1) another 200 g of water ... or (2) 100 g water and 100 g glycerin ... or (3) 200 g of glycerin.
If I wanted to have a water only recipe, obviously I'd use Option (1).
Either Option (2) or Option (3) will work great for a water-glycerin batch. I don't see a big difference between the two, except for cost.
I hope I'm not being too confusing here.