As EMI expands in higher education, academic writing has become a high-stakes site where students must simultaneously meet disciplinary demands and develop L2 writing competence, yet evidence on EMI-related writing gains remains mixed and highly task-dependent. This motivates longitudinal, multidimensional tracking of how interlanguage development is enacted across recurring genre “touchpoints” and time constraints, rather than inferred from single end-point writing samples. This study conceptualizes English-Medium Instruction (EMI) writing development as value-in-use enacted across a semester-long journey of genre touchpoints, with learners’ interlanguage functioning as the mechanism through which course resources are translated into observable outcomes. Using a within-subject longitudinal design, 80 EMI undergraduates completed five in-class timed writing tasks (three 20-minute touchpoints and two 45-minute essay touchpoints), yielding 400 texts. Texts were analyzed using a multidimensional outcome space (c-CAF): accuracy (error-free T-unit ratio; errors per 100 words), fluency (words per minute; total words), syntactic complexity (clauses per T-unit; mean length of T-unit), lexical diversity (MTLD), and cohesion (cohesive devices per 100 words). Mixed-effects and clustered inference tested (i) developmental change across the semester, (ii) systematic touchpoint and timing effects (20 vs. 45 minutes), and (iii) persistence versus genre sensitivity in recurrent error tendencies. Results showed robust semester gains in syntactic complexity, lexical diversity, and cohesion, while accuracy and writing rate did not improve monotonically. Timing and genre produced a consistent reconfiguration of the performance profile: 45-minute essays elicited substantially longer texts and higher complexity, lexical diversity, and cohesion, but lower writing rate and a small reduction in error-free production. Error-family analysis revealed persistent interlanguage pressure points (articles/determiners, lexical choice/word form, tense/aspect) alongside clear genre sensitivity: essays reduced article and preposition errors but increased lexical choice, tense/aspect, and sentence-boundary problems. The findings advance a journey-based account of EMI writing as value-in-use and offer actionable implications for designing touchpoint sequences that balance discourse expansion with targeted stabilization of persistent interlanguage constraints