There's no reason to believe the limited number of queens would have affected population sustainability.
In fact, we don't know that they didn't have queens before they started taking hosts and going mad.
The symbiote may hijack its host's reproductive system to do so (loosely similar to real life Goa'uld reproduction has always been described as "primarily asexual" but it is said queens have the ability to fertilize their own "eggs".
Most people don't even know these terms or think they're not possible or get them mixed up so I think you're getting hung up on the writing more than anything. The Goa'uld queens are shown to be able to choose the genes and memories of their offspring.
They have the ability to produce larvae that are genetic copies, but because of some genetic trait other than sex.
While queens (hermaphrodites) are capable of self-fertilization, non-queens (males) have been shown capable of inseminating a queen and passing on their genetic memories.
Goa'uld, aside from being endosymbionts, are aquatic predators that reproduce by spawning large amounts of offspring left to fend for themselves with only a few surviving to adulthood (traditionally labeled : the two sexes are males and hermaphrodites; in the absence of males the hermaphrodites may self-fertilize.
Where their reproductive cycle becomes bizarre is that only an extremely small percentage of the symbiote population are hermaphrodites, which doesn't make sense in asocial species.
Goa'uld hunger for power, and being able to control the continuation of their species is a great power. Such social conditions wouldn't have existed in their hostless days, as far as we know, and discussion of the original species that still existed on P3X-888 didn't shed any further light.
In general, we aren't told much about how the species operated during its evolutionary process.