The Advanced Benchmark Experiment (ABE) is a new step in benchmark experiments of fusion reactor technology aimed at examining the effects of non-homogeneities due to the discretization of the reactor blanket into breeder and coolant (confined within stainless tube), as well as to openings in the blanket for vacuum pumping, plasma heating, etc. The organization of the openings and any discretization significantly alter local nuclear parameters, particularly the local tritium breeding ratio (LTBR). Prior to designing an ABE, the practical limits of the quality of the experiment should be established and compared with the expected possibilities of numerical calculations. A study of the state of the art in LTBR measurements is presented. The neutron fluence is measured by the charged associated particle method with the use of ΔE- and E- silicon detectors. The tritium activity induced through nuclear transmutations of lithium isotopes is measured by a very advanced coincidence-anticoincidence system on direct mixtures of LiNO3 water solution and ATOM-LIGHT scintillator (the considered indicator mass is 0.1 g of LiNO3). The experimental results complemented by 3DAMC-VINIA code calculations reveal that in ABEs it is very realistic to expect 3.2% of maximal systematic error, and a statistical error ≅1.5% on LTBR measurements is achievable in most of a hollow sphere (R = 56 cm, r = 20 cm); this can be achieved with 9 days of an accelerator beam (Ed = 0.3 MeV, id = 0.5 mA).