Some companies run innovation through a single internal mechanism (Bosch's accelerator is the example in Explore/Exploit Dual Portfolio Management); others integrate multiple distinct levers under one overall strategy:
Ping An is presented as the case for the integrated approach, combining all four levers under one strategy while rising from roughly a top-500 company to roughly top 20-30 globally. The contrast with Bosch (internal-accelerator-only) suggests internal staged-funding programs like Bosch's are one lever among several a mature innovation strategy can combine, not a complete strategy on their own.
Osterwalder describes a portfolio-of-connected-programs model that extends the CVC/M&A/open-innovation/internal-team integration: leadership evidence-based-investment programs, hackathons, and a defined follow-on "vehicle" are explicitly linked into one pipeline, rather than each source of ideas running as its own disconnected activity.
The point of connecting hackathons specifically is to give hackathon outputs somewhere to go — see Hackathons as Innovation Theater for why disconnected hackathons fail, and Organizational Landing Pad for Scaling Validated Ideas for the landing-pad problem this pipeline is meant to solve.