First, the facts that shields recharge at least a few points fairly quickly means that a simple shields up equals no boarding type rule would make it almost impossible to get the timing right and effectively preclude real boarding.
Requiring a minimum of 10% or even 20% strength would mitigate this problem. I would expect this minimum to be with respect to the total shields of the class, not the surviving shield generators, so if you shoot out one or two before boarding it will be even less difficult - exactly the same as shooting out a few engines to improve the speed advantage of your boarding shuttle.
Second, if the rule is based on percentage of shields, even a small shield generator will prevent boarding from occurring, and a very small shield generator will actually be more effective at that, because the percentage will increase faster since the recharge rate is fixed and the capacity is small. In theory, the minimum size shield generator would be ideal for preventing boarding.
Is this really a problem? I'm not sure that using the minimum size (i.e., least efficient, because shields scale superlinearly with generator size) type of shield generator just to prevent boarding, which is usually not the major mode of ship-to-ship combat, would be a good plan even if it does work. And it doesn't work as well as you'd think, because shield recharge rate also scales with the size of the generator so while a weaker shield does recharge faster it is recharging significantly less actual shield points per increment than a large shield which makes it even worse for everything that is
not repulsing boarders. Basically, if you want to use very poor shields for the sole purpose of repelling boarders at the expense of ship-to-ship combat ability, I don't see why this is a bad thing to allow a player to choose?
As a result I think the best solution is that a boarding attempt functions somewhat like ramming, where the boarding ship would take damage and the target ships shield would be consumed. The question then is is it one to one shield to damage, or some ratio. I think this solution works well because it makes it harder to board ships with larger shields, and easier to board ships with smaller shields, reduces but does not eliminate the question of timing the boarding, and still incentivizes using weapons to reduce the shield before attempting to board. One possible caveat is that damage should probably be distributed across the boarding fleet evenly, so you can't just include a single large and heavily armored ship in your boarding party to absorb the damage.
Mechanically I do not like this idea, because to me it sounds like adding micromanagement and not simply another factor to consider. Boarding shuttles are not often very durable ships, so having them take damage from a boarding attempt is likely to cause internal damage or destruction quite often. This becomes problematic because it leaves you with not enough troop transport capacity to recover your troops, especially if you want to do more boarding attempts, scuttle the captured ship, etc. but your transports have been destroyed from bashing onto shields. I think it is better for the risks to be assumed by the boarding marines rather than the transport, as even if the marines take heavy losses they can be recovered after the battle is finished with minimal fuss (it is just giving one more order and sending the transport home).