Plant-based seafood alternatives could reduce pressure on marine ecosystems, but consumer acceptance remains limited by texture. Here we combine Texture Profile Analysis (TPA), consumer sensory evaluation, and data-driven modeling to ask whether large instrumental texture differences translate into perceptible sensory differences in breaded shrimp. TPA was performed on four breaded shrimp products, including one plant-based product and three conventional comparators. Consumer sensory evaluation used a within-subjects design with 107 participants comparing plant-based and conventional breaded shrimp through hedonic ratings, Just-About-Right attributes, Check-All-That-Apply descriptors, and purchase intent. Instrumental testing showed that plant-based shrimp was substantially harder, stiffer, and chewier than the conventional sensory comparator, with 1.6-fold greater hardness and 4.2-fold greater chewiness. In contrast, consumer evaluation revealed no significant differences between products for texture liking, overall liking, JAR firmness, JAR chewiness, CATA descriptors, or purchase intent after false-discovery-rate correction. These results suggest that the breaded coating masks mechanical differences in the protein core and that full TPA parity may not be required for consumer acceptance in breaded seafood formats. These paired mechanical and sensory data motivate an anisotropic data-driven model for plant-based seafood texture. Rather than treating texture as a scalar TPA target, the model will learn direction-dependent mechanical descriptors and map them to sensory outcomes, accounting for heterogeneous structures such as crispy coatings surrounding softer cores. This framework could help identify which mechanical differences are perceptually relevant and which are masked during oral processing, enabling texture optimization guided by consumer response rather than instrumental parity alone.