Ugh. Terrible interpretation. If that interpretation was correct, it would be the only effect in the game that let you rebuild magic items for free.
Talya, your question doesn't really have an answer, because how many "clutch" spells your character needs varies depending on campaign. For example, one campaign might not include any meaningful number of undead opponents, in which case all the undead-specific spells are of marginal use and can be sacrificed. In another, undead might be common but aberrations might never appear. One DM might resolve all social situations with dice checks, while another might go with how persuasively the player speaks -- in the second DM's campaign, spells that improve your social skills are useless. Bans and nerfs under various sets of house rules further complicate the situation. Further, we're comparing to schrodinger's wizard, who has an unknown number of spells known -- perhaps he knows 2/3 of the list, or perhaps he knows only his automatic spells plus whatever he can get from spellbooks looted from foes. Perhaps he has banned three schools, or maybe he's a generalist.
In the general case, though, the "tipping point" is the point at which the value of spontaneous casting and sorcerer class features exactly matches the value of faster progression, wizard class features, and slightly more spells known. I think you're probably going to hit that point when the sorcerer knows about 3/4 of the spells that the wizard does.