There's a lot to answer there. Weather and geography will play a key role for light vehicles, with tiny, non-flying vehicles like scooters always being better with batteries, while anything large SUV or greater will ultimately go fuel cells - the DOE earlier this year ~30% lower cost for SUVs and there are practical limits on larger systems like highway weight limits and the diminishing cost of balance of plant that make fuel cells make more sense.
In practical use you're at 55-65% energy efficiency, and experimental systems won't improve on that significantly. Generator systems can recover heat, so combined stationary systems are very high efficiency - serves a different use-case than batteries. The efficiency is scoped into expectation (1 kg is roughly equivalent to 4L of gas in an FCV - wheras with this co and Hydra I'd expect just over half that) and in general it's not as big a deal as the EV proponents make it out to be since batteries end up using a fair amount of energy car heating and/or battery cooling, while fuel cells can cold start or operate in really hot weather without losing capacity. FCVs can ramp to max power quickly, but peak power determines system size, which isn't ideal because repeat unit cost is relatively high; systems now address the acceleration issue by using larger batteries and supercaps - still a fraction of what a BEV would have - which also allows for regenerative braking.
Lifetime has improved significantly in the past decade or so due to additives (radical scrubbers, today typically Cerium nanoparticles), but that's the major area future material development is trying to address. For consumer automobiles the lifetime is fine (equivalent or greater than an 'ecoboost' engine I saw in one presentation), and for large vehicles stacks last up to an order of magnitude longer (FCVs are happiest when in consistent use) and replacement becomes a part of the maintenance cycle.
Hydrogen storage has a few solutions thanks to the composites work that Boeing pioneered with their systems - cost of the tank is a bit excessive but not seen as a scale problem per se. High pressure seems the way to go, and physisorbed or chemisorbed solutions will help improve the volumetric and energy efficiency of this - still active research, but a lot of candidates like cryo have been culled from the herd so even the focus is promising.
As far as electrolysis - again, we're close but with the low "traditional alkaline systems" are good enough for now, decently efficient and adaptable thanks to Zirfon separator (90's, a NL R&D centre licensed to AGFA) and are going in at the 100 MW scale in several countries in Europe - most of the new capacity going in is electrolysis, and we're at 5% renewables today up from 1% in 2012 according to the presentations I've seen - not bad for a $120B market - and use in fuel cells make the value proposition proportionally better because these require CO scrubbing, which isn't innately present in hydrogen out of electrolysis, and there's also a premium for pressure if you can get it - traditional systems operate at relatively low pressures, but the coming breakthrough is anion-exchange membrane based electrolyzers (AEM-WEs) - these fit the much smaller form factor of acidic PEM-based systems (~1/10th the active area) that can generate high-purity hydrogen at high pressure (I've seen claims up to 300 bar) without the requirements for Pt and Ti that drive more than half the system cost at scale.
