In haplodiploid sex-determination system, female/diploid individuals are developed from fertilized eggs by means of sexual reproduction, while male/haploid individuals are developed from unfertilized eggs by means of parthenogenesis.
So assume there is a highly intelligent haplodiploid species. All female/diploid individuals of this species have the humanoid appearance (even though they do not necessarily look like human females), while all male/haploid individuals are mere egg-sized "balls" that possess neither human form nor sentience. The male/haploid individuals are produced periodically in the way similar to menstruation (or hens laying eggs) and have a very short lifespan respectively. As a result, female/diploid individuals of this species would exchange the male/haploid "eggs" (or "testes") that they produce with each other when they need to reproduce female/diploid offsprings of next generation. Since those male/haploid "eggs" (or "testes") exist outside of the bodies of female/diploid individuals and are disposable by nature, this species is technically not hermaphroditic.
So my question is whether such a radical sexual dimorphism between female/diploid and male/haploid individuals is plausible or not. It seems that in most haplodiploid species like Hymenoptera (ants, bees, and wasps) the physical morphology of male/haploid drone individuals is not radically different from female/diploid individuals despite the fact that they are technically also sperm pockets exchanged between female/diploid queen individuals of different colonies. One possible reason why this may be implausible is that sexual maturity always takes a long time (maybe one decade or even more so) to reach for a highly intelligent humanoid species even though all male/haploid individuals are just insentient "eggs". In other words, it is simply too costly for female/diploid individuals to produce disposable "testes" on a regular basis.
Or maybe the reproductive mechanism of this species can actually work?
Update 1: Another possible issue may be that the radical sexual dimorphism between female/diploid and male/haploid individuals diminishes the main evolutionary advantage of haplodiploidy. It has been proposed that haplodiploidy may evolve as a mechanism to get rid of defective genes that are much more visible in haploid individuals. But if those male/haploid individuals are mere balls, then it would be nearly impossible for female/diploid individuals to observe the existence of those harmful mutations. That is to say, if male/haploid individuals are mere balls, then the risk of accumulating recessive traits (or inbreeding) would be greater for them.
Update 2: Since @Willk points out that this species is functionally equivalent to simultaneous hermaphrodites, the post now can boil down to the question whether or not it is evolutionarily plausible for a human-like species to evolve a way of fertilization without the involvement of penile objects.
PS: I prefer using terms like "diploid" and "haploid" to terms like "female" and "male" to avoid some unfortunate implications.