I'd point out that moisture, and in low enough quantities even minerals, may not add combustion energy but do serve as reaction mass. Unfortunately I suppose minerals would tend to produce abrasive ash which would be bad for the throat and nozzle. The low density of wood also makes it a good insulator so one might have low-temperature tolerant material wrapping the wood block for pressure containment; only as the wood layer gets to be very thin would enough heat leak out to soften or otherwise threaten the structure. But the basic problem of quality control remains, with the wood as pointed out tending to crack, creating hot spots on the pressure containment wrap and terminating the use of the wood early, and at an unpredictable moment, not to mention shedding large embers blocking the nozzle as also mentioned. To make the burn more predictable, one would have to break solid blocks of wood into thin layers for re-wrapping as plywood. It seems to me the main advantage of using wood for fuel would be the simplicity of just finding a suitably shaped block of natural wood, milling it to fit in the pressure containment jacket and drilling a central column for the initial burn, an advantage completely discarded if we have to build the wood up as plywood or particle board. If we have to build it up we might as well use paraffin or some such instead. And the insulation factor is a bit of a red herring; hybrid solid fuel elements insulate by ablation, the outer layers are kept cool enough since the inner layer that is burning is evaporating. Indeed, one limit on development of hybrid rockets is that it is not easy to get the solid fuel to evaporate rapidly enough to deliver thrusts comparable to traditional solid propellants that include the oxidant!
Genetically engineered "stage trees" a la Niven of course are not hybrid since the oxidant is mixed into the fuel to make a more traditional solid propellent, that cannot be controlled. Presumably however the stage trees are very sophisticated in the respect of layering their combustion elements for a prescribed burn, and having both very high thrust (to minimize gravity loss during ascent) and very low propellent/"dry" mass fraction, achieved I suppose by a combination of very light yet strong outer skin (something like spiderweb or Kevlar strong--Kevlar also can tolerate fairly high temperatures) and strong but combustable cell structure. I figure the mass ratio of propellent to spent mass, of which the interstellar "seed" would only be a portion, must be well over 1000. We can't do that with human fabricated rockets but I suppose it might be possible for something biologically grown to achieve it. It has to, to achieve total delta-V from the surface of 20 km/sec or more to achieve system escape velocity and thus send its seed at a decent speed between stars. Once launched I suppose it deploys some kind of solar sail to maneuver in close encounters with stars, to either aim for another candidate star or maneuver to collide with a suitable life-bearing planet. Depending on how big this "seed" needs to be to survive hundreds of thousands to millions of years coasting from star to star--I'm guessing a metric ton, give or take an order of magnitude--I'd suppose the tree itself--the actual rocket trunk, not other vegetation with leaves and roots to obtain the necessary energy and nutrition to grow the trunk and seed--would have to mass from a thousand to a hundred thousand tonnes!
And it would have to "know" the delta-V needed to send its seed on from the host planet. Fortunately this information is available from its entry into the host star system, its detection of and approach to the host planet, and refined by observation of planetary atmospheric entry details. This can all be integrated into a simply benchmark number that can be encoded in genetic material and guide its detailed growth strategy, aiming for a certain size to accomplish another interstellar launch. The tree also has to pick a target star, in the plane of its host system's ecliptic, and time its launch to aim for it. Given the sort of optics it would need for interstellar navigation I suppose this is fairly easy.
So yeah, stage trees could be a thing, though I don't think much of their chances of evolving by sheer chance. Niven's aren't supposed to have randomly evolved though; they are offspring of genetically engineered products meant to provide cheap launch capability to the "Slaver" species aka Thrintun, and their engineers, the Tnuctipun slave/rebel species, may have secretly installed modules predisposing them to interstellar spread and survival, now honed by a billion years of natural selection after the Tnuctipun took down the Slaver empire in a mutually destructive rebellion.