Despite contradicting evidence that open access (OA) articles might have greater citation advantage, there is less case studies in developing countries showing whether their global publication availability pattern advantages scientific impact metrics. Also, by addition of altmetrics to the world scientific evaluation system it is less known how different research access channels such as OA publishers, PubMed database and arXiv repository help altmetric indicators. Therefore, this paper investigates the case of WoS publications of Iran (2001-2012) for impact of mentioned publication availability models on citation, Mendeley readership, and tweet counts across four broader disciplines. Findings on 98,453 articles show that gold OA papers (5%) do not benefit significantly more metric counts, except in tweets linking to OA medical publications. Articles in PubMed Central (3%) significantly advantage the three investigated metrics, whereas arXiv preprints (2%) had higher readership advantage only. Different from PubMed publications, tweets to OA medical research were not significantly correlated with citations, suggesting their social impact rather than scientific. Additionally, OA publications are not significantly read by Mendeley users in developing countries, but developed ones, only in life science and biomedicine. Therefore, repository availability appears to be highly impactful in terms of citation and readership, whereas OA publications tend to receive rather high social impact through tweets.