Thanks for that feedback :) I suggested it because OP asked for suggestions and thoughts for their new language. I'm kind of peeved at the downvotes because at a minimum it's a reasonable idea to look into - as you've just confirmed, saying you've picked it as the best solution.
I thought it's worth considering as the syntax solution for a new language! It's fully orthogonal and removes all style decisions. The compiler never has to guess. That actually means the IDE could be fully sure of it in every case and could let you adjust the tabs, as in this example I put together (using excel and a C program):
https://imgur.com/a/s7zQMpw
----
I chose my suggestion based on orthogonality. Originally I started suggesting 2-space tabs for OP's new languages (as a syntax feature), since they're a pretty common style convention used by many in the C world, so not totally absurd.
But that is not fully orthogonal and hard to automate: if the compiler encounters 3 spaces, what is that supposed to be? Automatically a syntax error?
How many indents is 6 spaces again? -- when you use 2-space tabs the answer is 3
//Guess below line's indent level (2-space indents)
x
With two-space tabs, the answer is 3:
//This is indented how many?
x
a
b
c
x
but when you use 1-space tabs the same is:
//Guess below line's indent level? (1-space indents)
x
//Answer:
x
a
b
c
x
You can just count, and it is visually also very clear.
As mentioned your IDE can treat leading spaces as indents. (Visually indent them.) It's a 1:1 correspondence.
Anyway just my humble suggestion for tomcam's new language, who solicited ideas with "Any ideas?". I gave actual code snippets in my original comment. :)