The site's wording asks for the possible values of $f(n)$ for every $n$, as Grünbaum's question is stated in Erdős's 1972 paper (p. 23) and in Salamon and Erdős [ErSa88]. The site labels the problem SOLVED and its commentary says "Solved (for all sufficiently large $n$) completely by Erdős and Salamon [ErSa88]; the full description is too complicated to be given here". The curator's reading is the question for all sufficiently large $n$, and the corrected Statement adds those words; nothing else changes. The evidence for the discrepancy is the paper itself. Salamon and Erdős write (p. 137) that their formulas give "a complete answer to Grünbaum's problem for $n\ge n^*$" but leave "the problem for $n<n^*$ open", that this case "requires a detailed analysis of the lower end of the high $k$ bands and appears to be difficult", and that "the size of $n^*$ is unknown but it is likely to be small"; their figure 5 shows, for $n\le12$, values at the lower end of the continuum below $\binom n2$ that the large-$n$ formulas omit. The answer under each reading: the corrected Statement is answered by [ErSa88], which determines the set of values of $f(n)$ for every $n\ge n^*$ band by band, with Erdős's 1972 theorem supplying the values above $c_1n^{3/2}$ and the paper fixing the best constant $c=1$ for the lower end of the continuum; the question for every $n$, as printed, is open for $n<n^*$, a threshold the paper does not compute, so no finite list of exceptional $n$ is on record. No result about the site's wording beyond these two papers is known here. [Er85], the site's source key, records Grünbaum's question of which values the number of lines can take, and reports without proof that every value above $cn^{3/2}$ occurs except $\binom n2-1$ and $\binom n2-3$ (pp. 2-3, source page).