"The chair was ugly but uncomfortable."
is an example of a paraprosdokian, that is,
a figure of speech in which the latter part of a sentence, phrase, or larger discourse is surprising or unexpected in a way that causes the reader or listener to reframe or reinterpret the first part.
This definition mentions reinterpretation of an earlier part of the sentence, which some might claim is absent in your example. I think that there is reinterpretation present in your example, namely, but must be reinterpreted without it's contrastive connotation. I thus think paraprosdokian (as defined above) works to describe the example.
In short, use of but raises an expectation of contrast, which is then defeated by the latter part of the sentence. This defeat is humorous, as are many examples of raising and then violating an expectation—a fact observed by Aristotle in Rhetoric and Cicero in On the Orator (here).
There is also a broader definition of paraprosdokian as
A sentence in which the latter half takes an unexpected turn.
On this broader definition, the above example is uncontroversially a paraprosdokian. After all, the final word is quite unexpected.
I thus think paraprosdokian works to describe the example on both the stricter and broader construals of what it means to be a paraprosdokian. Whether or not there is a more specific name for this play on but (or the general device of raising an expectation and then defeating it) remains to be seen.
For those nay-sayers still unconvinced with my classification of the example as a paraprosdokian on its strict construal, contrast the above with the following also slightly humorous example:
"Mary walked to the cliff and jumped, but not in that order."
As Paul Grice noted in the 60s, and is commonly used with what might be called order connotation, that is, event descriptions conjoined with and are assumed by hearers to be in their temporal or causal order. But sentences like the one above have an additional clause tacked on which subverts or defeats this connotation. This additional material forces a reinterpretation of the earlier part of the sentence. Namely, and must be reinterpreted without its order connotation. I thus classify this sentence as a paraprosdokian.
I see the original sentence and the second sentence as analogous in the relevant respects, making them both paraprosdokians.
Given these arguments, I think the burden of proof has shifted onto those who would not classify these as paraprosdokians.