Correction and strengthening of ``How large are the level sets of the Takagi function?'' (Q466844)

Let \(T: [0,1]\to\mathbb{R}\) be the Takagi's continuous nowhere differentiable function: \[ T(x)=\sum_{n=0}^{\infty}\frac{1}{2^n}\phi(2^nx), \] where \(\phi(x)\) is the distance from \(x\) to the nearst integer. It is well-known that the range of \(T\) is equal to \([0,2/3]\). For \(y\in\mathbb{R}\) let \(L(y)=T^{-1}(y)=\{ x\in [0,1]: T(x)=y\}\). The author proves that the set \(S^{\mathrm{loc,uc}}_\infty\) of all \(y\in\mathbb{R}\) for which \(L(y)\) contains uncountably many distinct local level sets, is residual in the range of \(T\). This corrects an error and strengthens the main results from an earlier paper by the same author [Monatsh. Math. 167, No. 3--4, 311--331 (2012; Zbl 1256.26002)].