For this answer, I'm going to try to focus on broad-strokes balance issues, rather than particular broken combos.
The classes will feel much more similar.
Druids and clerics are both prepared casters who use WIS as their casting stat, but have very different spell lists. If you remove the spell list differences, these classes' abilities will become very similar. Balance-wise, this means everyone can optimize based on which class has the best non-spellcasting features, instead of having to factor in the spell list.
Classes won't have distinct weaknesses.
It's by design that wizards are very bad at healing, and clerics are bad at blowing things up. This is a balancing factor that's meant to counteract the 3e-era problem of "the best party is always four wizards" and incentivize playing a mixture of classes. In base 5e, four wizards will have great DPS and crowd control, but no healing and very little defense. With this change, that's no longer a concern.
Prepared casters are much more versatile.
Spontaneous casters won't get as much of a benefit from this, since even if they have access to an order of magnitude more spells, they can't learn any more. Prepared casters, on the other hand, can take advantage of this new panoply of options at every rest. (And of course classes without spells get no benefit at all: fighters and rogues get left out in the cold.)
Which also means…
Decision paralysis gets worse.
A prepared caster like a cleric now has the option of preparing any spell in the entire game at a long rest. That's a lot of options! The player has gained a whole bunch of extra power, but only if they can OOC think through all those options…and this comes up anew every day. This puts more burden on the player's OOC planning and decision-making skills to take advantage of the expanded list.
A lot of choices get taken away.
Spontaneous casters, on the other hand, still have a very limited number of spells known. With all the spells in the entire game to choose from, it's likely all of those choices will get filled in by the objectively best spells.
Currently, a ranger might take a couple of "mandatory" spells that are the best on their list, and then fill out their other choices with fun or flavorful effects that aren't as potent. But with this change, the opportunity cost for choosing those fun spells becomes much higher—because you're taking them instead of the best spells from the wizard list, the cleric list, and so on. While prepared casters may end up with too many choices to make, spontaneous casters are likely to end up with too few!