##
Yes — your food-chemistry intuition is essentially right. A Frosty is much closer to a carefully engineered, low-fat frozen dairy suspension than to ordinary hard ice cream.
The formulation varies somewhat by market, but Wendy’s published ingredient lists are revealing. The current UK chocolate Frosty, for example, is primarily water, sugar, skim-milk powder, a relatively modest amount of fat, whey, cocoa, emulsifier, and hydrocolloid thickeners (guar gum and xanthan gum). A 12-oz serving has only 6.9 g fat. (Wendy's) U.S. formulations reported by Wendy’s have similarly used milk/cream plus guar gum, cellulose gum, carrageenan, and mono- and diglycerides. (Reddit)
So what holds a Frosty up?
Think of four structural elements operating simultaneously:
1. Ice crystals are doing a tremendous amount of the work.
At its dispensing temperature—reported around 19–21°F (−7 to −6°C)—a substantial fraction of the water is frozen. (Allrecipes) Those microscopic ice crystals occupy volume and jam against one another, giving the Frosty its spoonable yield stress.
This explains exactly what you've noticed: warm it only a few degrees and you don't merely make the same substance slightly warmer—you melt part of its structural framework. The solid ice fraction drops rapidly, viscosity collapses, and suddenly you've got something remarkably close to chocolate milk.
2. Sugar controls how much ice exists.
Sugar depresses the freezing point, so at −6°C not all the water freezes. You therefore have ice crystals suspended in a concentrated, sugary aqueous phase rather than a solid block of flavored ice.
This is one reason the temperature window is so important. A few degrees colder → too much ice, unpleasantly hard/icy. A few degrees warmer → too little ice, soupy.
3. Hydrocolloids make the unfrozen liquid much more viscous.
Guar gum, cellulose gum, carrageenan, xanthan, etc., bind/immobilize water and dramatically increase viscosity at tiny concentrations. They also inhibit ice-crystal growth and help maintain a smooth texture.
But importantly, they aren't sufficient to make a room-temperature Frosty into Frosty. Once most of the ice melts, you've lost the principal particulate scaffold. Hence the striking transition you've observed.
4. Fat + milk proteins + emulsifiers provide creaminess.
There is a dispersed fat phase, but considerably less than in premium ice cream. Milk proteins and emulsifiers help stabilize the fat/water/air interfaces and produce the perception of creaminess without requiring 12–16% butterfat.
There's also air (overrun) incorporated during freezing. So microscopically you can imagine:
ice crystals + air bubbles + tiny fat droplets → packed into a thickened sugar/milk/water continuous phase.
That's the Frosty.
And this gives a particularly nice explanation for why it behaves differently from, say, Häagen-Dazs. Premium ice cream has a substantial fat structure that continues contributing body as it warms. A Frosty is disproportionately dependent upon its frozen-water structure. Melt the ice and its apparent "creaminess" rather abruptly reveals itself to be a comparatively lean dairy liquid.
So your observation — "very cold → remarkably thick; melts a little → rapidly becomes chocolate milk" — is actually a rather good informal experiment demonstrating that ice volume fraction, rather than fat, is the dominant structural variable.
##
I think the air bubbles are extremely tiny, much smaller than I’m used to seeing an ice cream.
##
I think you're right. And that is probably an important part of the distinctive Frosty texture.
In ordinary soft-serve or ice cream, the incorporated air bubbles can be on the order of tens to hundreds of microns, and their presence contributes to the fluffy, creamy appearance. A Frosty seems much more dense and extremely fine-grained—you don't perceive obvious foam cells.
That can result from the way it is frozen and dispensed. The combination of continuous agitation/shear, stabilizers, milk proteins and emulsifiers can produce very finely divided air bubbles, while simultaneously keeping the ice crystals small. The sensory result is interesting: rather than "fluffy soft serve," it feels almost like an extraordinarily cold, thick liquid.
And your melting observation supports that interpretation. If a large part of its apparent volume and body came from a robust foam structure, you'd expect the melted product to retain more mousse-like or foamy character. Instead, a melted Frosty rather quickly becomes a fairly ordinary thin dairy beverage.
So I'd picture the fresh Frosty microstructure approximately as:
lots of tiny ice crystals + very small air bubbles + microscopic fat droplets, all crowded together in a sugar/milk/protein/gum solution.
The ice crystals provide much of the mechanical stiffness; the gums make the liquid between them viscous; the tiny air bubbles soften the perception of the ice; and the emulsified fat and milk proteins supply creaminess.
One correction to my previous answer: I wouldn't confidently say that Frosty's air bubbles are known to be smaller than conventional ice cream without microscopy or published particle-size measurements. I don't know of Wendy's having published those data. But your sensory observation is quite consistent with a low-overrun and/or very finely aerated frozen product, and the absence of perceptible bubbles is certainly striking.
##
The cheap ice cream of the Midwest of the 1960s melted into and almost foamy mixture. I assumed it was revealing the relatively large air bubbles that had been blown into it.
##
Yes. That is very likely what you were seeing.
A lot of inexpensive mid-century ice cream was made with high overrun—a great deal of air whipped into a relatively inexpensive dairy mix. If the overrun were 100%, for example, a gallon of liquid mix became roughly two gallons of ice cream. The air wasn't necessarily in enormous visible bubbles while frozen, but the dispersed gas phase occupied a very large fraction of the product.
When it melted, several things happened:
The ice-crystal framework disappeared, releasing liquid water.
The air remained temporarily trapped by milk proteins, emulsifiers and stabilizers, producing that peculiar light, foamy melt.
Eventually the bubbles coalesced and escaped, leaving considerably less liquid than the apparent volume of the original ice cream suggested.
So the melt itself was almost a crude demonstration of the formulation. Cheap, high-overrun ice cream → foamy, aerated melt. Dense premium ice cream → relatively rich, creamy melt. Frosty → comparatively thin, milk-like melt.
And that last contrast is why your observation about Frosty is interesting. It doesn't seem to reveal a huge hidden foam structure as it melts. Its remarkable solidity when freshly dispensed seems disproportionately attributable to a very high concentration of tiny ice crystals at a carefully controlled temperature, rather than to either lots of butterfat or lots of air.
In fact, if you left equal apparent volumes of 1960s bargain ice cream and Frosty to melt completely, I'd expect their respective puddles to tell you quite a lot about how differently the manufacturers achieved the original illusion of "creaminess."
