If the stated goal is to unify, then it must be possible (to a significant degree) for people who love Vancian to have it, and those who loathe it to avoid it. At the very, very least this means they'll try to make sure a person can play a character that matches their preference. For a great many people that will be enough. The next toughest crowd are those who won't play in a group where anyone, even monsters or other players, might possibly use the loathed rules. For this demographic, playing D&D then depends on finding a group that is willing to not use the loathed rule at all. This is certainly possible for many of the most polarizing rules, except in those areas where any gaming group is a rare find. Still, for the vast majority, it will probably be enough. Finally, there is the absolute toughest group of people who might in principle play D&D: those who won't play with a system that supports the loathed rule, even if it is actually avoided like the plague in some group. I have to believe there are very few people who would avoid a game table that forbids Vancian casting principally because the system allows Vancian casting. It'd be like me hating Incarnum so much that I wouldn't play at a 3.5 game where it will never be used. I won't say they don't exist, but as a marketable segment? No.
Personally, I think there is room for even building Vancian and non-Vancian casting into the same basic structure, without entirely separate systems (not that I'd object to that either). Start with a spell point system, for example, and let a character "memorize" spells that removes from the spell point pool, but adds bonus "effective" spell points that can only be used to cast the memorized spell. A caster can then spend all their spell points in this fashion every day, and totally emulate Vancian casting. For people who are bothered that their character could, in principle, not use this Vancian method, simply make a little feature they can take that says this caster must spend all its points every on memorization. Other people could remain totally spontaneous, as still others could find their own balance. That is how the Wizard in my homebrew system works, at least. (The other details of how spellcasting works in my game make sure that the Wizard doesn't end up with so many extra spell points they can spam with impunity.)
Personally, the improvement (IMHO, of course) I most hope for from spells in 5e is continuity. In all existing editions of D&D it bothers me that something like "Delayed blast fireball" is separate from "Fireball", and can even be learned without the latter. Metamagic in 3/3.5 eased my pain somewhat, but it was clunky for all but the most basic improvements. And the replacement of powers in 4e irked me even more. I would much rather that spells were base spells plus small trees of upgrades. It is scalable (the upgrades), avoids pointless repetition (a robust base spell list), and flexible (not every upgrade needs to be universal). And assuming learning new spells or upgrading existing ones comes from the same basic "pool" of character resources, whether a caster wants to master a very small set or dabble in lots of base spells could be their choice. Finally, if you want to emulate the old-style of spells (treating a base spell + given upgrades as a unique spell), it is a paragraph of rules at most. To me that meets what appears to be the relevant design goals of the 5e team.
