Now you are asking the real questions, and a damn good one too. In this case, you have to remember that the spell slot and metamagic was so heavily balanced away from spell slot creation that anything that was not metamagic, felt like a waste. Ergo, to begin with, there was no real balance. Rather a scale with a big fat note of 'Metamagic' holding it down.
Now that I have had time to do some math, I have to ask how you reached this conclusion. Establishing a concrete baseline value for a sorcery point is always going to be tricky matter, but there are ways to approach it.
As I am playing and deciding whether or not to empower a spell, the comparison I make is to quickened spell. In its most common usage, quickened allows the sorcerer to cast a cantrip on the present turn, and, if that cantrip is a damaging one, it is most likely to be fire bolt. Because quickened costs two sorcery points, its value gained per point, in present-turn damage, is half the average damage of fire bolt: 2.75 at first level, 5.5 at fifth, 8.25 at eleventh, and 11 at seventeenth. (Note I am discounting things like chance to hit and damage type.) I like this approach because it scales the value of sorcery points with player level, so that I am less likely to frivolously empower spells when their damage is still reasonably near the expected mean.
Speaking of empowered spell, that metamagic provides an alternative means of valuing a sorcery point. If you use it to reroll the damage of the highest raw-damage single-target spell at each spell level (assuming for convenience a +5 Cha mod, and excluding scorching ray because of its awkward interaction with the metamagic), you find that it adds 3 present-turn damage to a 1st-level chromatic orb or catapult, 4 to the same spell from a 2nd-level slot, a little more than 5.8 to a fireball or lightning bolt at 3rd, almost 6.9 to vitriolic sphere at 4th (rerolling
both instances of damage, which is usually not actually a good play using the quickened test), a little over 7.2 to vitriolic sphere at 5th (same deal), more than 6.8 to disintegrate at 6th, almost 7.2 to the same spell at 7th, approaching 7.5 to the same at 8th, and 9th is difficult to calculate because of the huge number of dice of meteor swarm. It would, I guess, approach 12.5 in theory. All of these numbers are only considering a single-target, for the record.
One could perform a similar check with heightened spell, though that would require making some assumptions about save chances. Roughing it out with disintegrate and an optimal 50% save chance, it would add 6.25 present-turn damage at 6th spell level and 8 at 8th. Heightened spell is basically only worth using with disintegrate and plane shift, because most other save spells do half damage on a success.
Similarly with twinned spell, the calculation is possible but a little more difficult because of the requirement of two targets. Twinning a 1st-level chromatic orb would add 13.5 damage for one sorcery point, 2nd level would add 9 damage per point, and then things get blurry because the best damage spells until disintegrate cannot be twinned. Twinning a 6th-level disintegrate adds 12.5 present-turn damage per sorcery point, and the number drops to 12 by 8th spell level. But again, this is complicated by twinned spell's restrictions and split damage.
I would thus value the quickened and empowered estimates more than the others when estimating the value of a sorcery point. It starts around 3 present-turn damage, rises to 5.5 or so at 5th player level, 6.8-8.25 at 11th, and perhaps as high as 11-12.5 at 17th. The average for the entire range of levels, and especially those where most groups play most of the time, is probably in the 6-7 range. How does creating spell slots compare?
If you spend two sorcery points to create a 1st-level slot to cast chromatic orb, you gain 6.75 present-turn damage per point. At 2nd level, you get 6. At 3rd, 5.6, though note that this is the point where the large AoEs and powerful buffs start taking over as the top spells, so single-target damage might be a little misleading. Spending six sorcery points for a 4th-level slot to cast vitriolic sphere, you get 6.25 damage per point. And with a 5th-level slot, you get a little more than 6 damage per point.
I conclude that creating spell slots is a better use of sorcery points at lower levels, they're about even through the meat of the game, and metamagic becomes clearly better at the higher levels. This is also tempered by the enduring value of low-level slots; things like absorb elements, mage armor, shield, mirror image, and misty step scale with
enemy damage and chance to hit.
Thus, I have to disagree with you that creating spell slots is not balanced against metamagic.
In the current economy listed, it makes 1st and 2nd slot creation feel like a metamagic option. Because they have that approximate cost. While anything of 3rd and above, is one point cheaper, aside from 5th level spells, which has the same cost as originally. As such, due to the fact that most level one slots are utility slots, and second level spells are not the go-to slot for high damage or battlefield domination, this is not so bad indeed. Also, a sorcerer has so few known spells that their 1st and 2nd level spells are limited in selection if you want a versatile list later.
In the re-balanced economy, it sort of averages out. Because: "Hmm, sure, I could get a third level slot, but I would really like to augment x spell. Perhaps I should just sacrifice a 3rd level spells for points, if I suddenly need a third level slot, I can convert them back mid combat as a bonus action." Because as you might notice, the conversion for 3rd and 4th level spells, is in fact 1:1 Giving you a lot of flexibility.
For a sorcerer, this is not really breaking anything whatsoever, as the very limited spell choices you have means that you do not have an incredible library of spells for each level to choose from. Each spell slot has a limited amount of utility without metamagic.
The type of "flexibility" that you see this giving the sorcerer looks a lot like an undo button, which I would view as antithetical to the basic conceit of role-playing games that the player's choices matter. But does it break anything?
Well, that odd situation where sorcerers can create 4th- and 5th-level spell slots before they would normally have them from character progression? You just extended it down to 2nd- and 3rd-level slots. A 5th-level sorcerer could have a 4th-level slot
two levels before normal progression would give it to him or her. And then he or she could crunch a 3rd-level slot and a 1st-level slot to get another 4th. And again. It would make sorcerer the exclusive best blaster class until at least ninth level, and even after that it would have a considerable advantage. I would call it broken.
I'll be brief about multiclassing, since I have other things to do. For a non-caster, three levels in sorcerer would grant four 1st-level slots, two 2nd, and three sorcery points. The points could be converted to 1st-level slots, and then both 2nd-level slots broken down and also converted, giving the character up to eleven casts of absorb elements, shield, charm person, etc., per day, compared to seven RAW. A warlock could gain that many slots for his or her non-scaling spells and save the pact-magic slots for the scaling stuff. A paladin could smite for days. And that's on top of the normal advantages of multiclassing with sorc--Con-save proficiency, subclass features, metamagic, social skills. It still looks broken to me.
CapnZapp:
The fact that sorcerers should just get a redesign moment like they did for the ranger, in preparation for the next official book release is something I am entirely in agreement with.
These threads are made and in the end most to everyone agrees that sorcerer has some sort of issue, which makes the the situation rather plain.
It's an echo chamber. It's the same small group of people making the same half-considered statements and nodding in agreement with each other, in thread after thread month after month.