Evidence: Well-supported for the cited papers. GHz and “wafer-scale” in headlines are not a radio product.
Process — copper CVD
Li, Cai, An, Kim, Nah, Yang, Piner, Velamakanni, Jung, Tutuc, Banerjee, Colombo, Ruoff, Science 324, 1312 (2009). Large-area graphene grown by CVD on copper foils, mostly monolayer, transferable. This is the process later fabs and foundries keep trying to industrialize. It does not erase wrinkles, grain boundaries, or transfer polymer residue.
Competing routes still running: Ni CVD (often few-layer), SiC epitaxy, oxide/reduction, liquid-phase platelets.
Devices
IBM and academic groups already report graphene FETs operating at GHz frequencies on small gates (Lin et al. lineage; a 2009-era Nano Letters / arXiv thread). Cut-off frequency is a small-signal RF figure. It is not a CMOS logic roadmap and it is not power gain at system level.
Industry / media
Specialist outlets (Graphene-Info from this period) begin weekly vendor roundups. That creates a public record and a selection bias toward whoever issued a PDF. Platelet and ink companies are louder than film companies because powder is cheaper to announce.
Named early industrial interest (US/Asia filings and labs, not a complete census): IBM (SiC and later CVD devices), Samsung and other display groups on CVD film, platelet vendors that will later be cited as Angstron / Vorbeck-class firms. Put a firm on companies only with a dateable disclosure.
Claims vs open
Claimed: transparent electrodes replace ITO; graphene RF replaces III–V. Open: yield, mobility after transfer, contacts, and a band gap if you wanted a switch. Those remain open into the Flagship years.