There is no single form of a lisp; it's just a generic term that captures all sorts of messed-up ways of pronouncing sibilant sounds like "s" or "z". So just because gays don't produce a sound exactly like "th", but something close to it (a "hyper-corrected" form), is just trying to wave the problem away. They obviously have some very distinctive error in sibilant pronunciation, and to the ear of the average person the closest thing it sounds like is "th". Hence the common phrase "gay lisp".
Moreover, I've heard the gay lisp even among Spanish-speakers, whether from Spain or from Latin America, and in a handful of East Asian languages like Mandarin Chinese, Korean, and Japanese. So it's something that strikes them regardless of their native language. It is not a language-specific thing like choice of slang, but a common involuntary disruption to queer speech.
Here are some clues that the gay lisp is another part of the larger pattern of gays being stunted in childhood:
As a functional speech disorder, lisping has no clear known cause. It is often referred to as a speech delay of unknown origin. . . .
Lisping is also associated with immature development. Some children will adopt a lisp as a means of gaining attention. Other children will begin to lisp after they have experienced unusual stress or trauma. This behavior is part of a regression into a more secure period and can include other types of regressive behaviors such as bed wetting or wanting to sleep with the light on in the bedroom.
First, it is a delay, meaning stunted growth along the natural developmental path. It's not any old error like pronouncing "b" like "k". It's the kind of error that a stunted child makes, although most normal children will outgrow it.
And second, a child may speak with a lisp not because they got stuck in an immature state, but are regressing there after already having reached the intended, mature state of pronunciation. Regressing in order to gain attention -- reminds you of every faggot drama queen you've ever run into, doesn't it?
The review says that lisping tends to disappear for most children after a trouble period of ages 4 to 8. That reinforces that claim I've been making throughout this series of investigations into gay Peter Pan-isms that they're stunted after toddlerhood but before older childhood when they're on the cusp of puberty. They are in that resolutely "girls are yucky" phase of elementary school before 4th or 5th grade, when boys start to mature more socially, emotionally, and physically.
Finally, lisping is neither an exaggerated masculine or exaggerated feminine way of speaking, so we can rule out the popular yet wrong-minded views of male homosexuality that see it as hyper-masculinization or feminization. Females are more neotenous, so it is easy to confuse feminization with infantilization. But then along come cases like the gay lisp or the lack of a nurturing instinct that conclusively point to stunted development as the root cause.