Even if measured IQ is highly variable before 3rd grade (I'd have to review the research on this), this is not a strong argument against identifying students in K, 1, and 2, as long as you also identify students in 3rd and later grades. Having multiple entry points should not be problematic, but whether students identified as gifted in 1st grade should also be assessed in later years and possibly removed from a gifted program is a difficult question. The Johns Hopkins talent search requires students who qualify by doing well on the SCAT in elementary school to do well on the SAT, ACT, Advanced SCAT, or Spatial Test Battery in 7th grade to maintain eligibility for various programs.
Agree on multiple entry points. That should be a requirement of any successful gifted program... not only because children within your own community could have previously been missed, but also because community members come and go, and an incoming child may have come from an area where gifted services are lacking, and no prior opportunities for identification had been available.
Disagree on the difficulty of the question of reassessment. As you said, Johns Hopkins requires periodic reinforcement of qualifications. My state gifted program has a similar provision, in which G/T students have an annual end-of-year IEP review with parents and faculty, and a requirement to re-evaluate students for eligibility at least every three years, though that requirement can be waived if both parents and faculty agree it's not necessary. In practice, this means that if every evidence is that the student is not falling behind or seeming overwhelmed in G/T, they stay in the program without re-evaluation. So there's two examples of the question being answered rather easily.
