The accurate estimation of reactor physics parameters related to the presence of cruciform absorber blades in boiling water reactors (BWRs) is important for safety assessment and for achieving flexible operation during the cycle. Characteristics that are affected significantly include distributions of the total fission (Ftot) and 238U capture (C8) rates for controlled regions. Representative experimental investigations have been performed in the framework of the LWR-PROTEUS programme. In particular, the LWRPROTEUS I-2A experiments concerned the neutronics characterisation of a SVEA-96+ BWR assembly controlled with a hafnium (Hf) blade under full-density water moderation conditions. The current paper presents the comparisons of the measured Ftot and C8 pin-wise distributions with a variety of stochastic and deterministic calculations: (a) MCNPX2.5 using recent nuclear data libraries (JEFF-3.1, ENDF/BVI.8, and JENDL-3.3), (b) PHOENIX4 using ENDF/B-VI.3, (c) BOXER using JEF-1, (d) CASM04 using JEF-2.2, and (e) HELIOS 1.6 using ENDF/B-VI.l. The calculation/experiment comparisons show standard deviations from 1.2% (MCNPX2.5) up to 1.9% (BOXER) for the prediction of the Ftot distribution, the highest individual discrepancy (7.6% with BOXER) being seen close to the "Hf-vertex." The C8 comparisons show systematically better agreement than those of Ftot, the lowest standard deviations being 1.0% (BOXER) and the highest only 1.4% (HELIOS). In addition, sensitivity studies highlight the greater importance of modelling aspects, compared with that of nuclear data libraries, for the achievement of satisfactory and validated Ftot and C8 predictions.