do you want the long or short answer? do you even know how a laser works? I'm beginning to think you might not. but in short, you in theory could just line them up and get a little coherent light, but real world losses wouldn't allow that. (this is called single-pass gain)
Think of a HeNe...the very model of how a laser works. a gas tube (the lasing medium) with a mirror on each end... a high reflector (HR) (99.99% reflective) and an output coupler (OC) (usually around 98.5-99% reflective). when you excite the gas in some manner, it gives off coherent light, this is then trapped between these mirrors, creating a pool of light at that particular wavelength desired, based on the properties of the mirror. however, one mirror is slightly less reflective (the OC) than the other, so light eventually starts to 'leak' out. that's how you get your beam.
This principle is called a Fabry-Perot or FP style cavity, and it is the most basic form of creating a laser.
your pointer is made based on this process. it starts with an 808 diode. this light then passes through the Nd:YAG which causes the Nd atoms stuck in the crystal lattice to fluoresce at another wavelength (946) however single pass isn't going to give you much as it passes through, and will result in mostly waste, so a mirror is placed at each end using the same principle as above to gather enough 946 to be useful. this is then allowed to pass through the LBO for its doubling. however again, we need more than one pass. so putting the OC before the LBO won't work. so it's put after, but we also need it to reflect 473. so the mirrors are then made to reflect that as well. and the LBO is coated AR to allow them to pass through with less reflections, allowing as much to be doubled as possible. the remaining 473 is allowed out the OC to be coupled by a lens and used by you, while an IR filter, catches the last stray IR photons.
however this is how it is done on an optical table. but to save space and cost since the pointer is so small, these mirrors are often coatings on the surfaces of the crystals, instead of separately mounted parts. this guarantees the surfaces are always parallel as long as the cavity is aligned properly (minus thermal expansion) with the OC. and saves space, as well as cost as you don't have to fabricate actual mirror substrates. they just apply them directly to the crystals. it also makes it simpler (relatively) to make.
you can use this pic for reference:
mirrors would be:
Nd:YAG: HR on first surface, AR last surface
LBO: (AR both surfaces)
OC lens is well...the OC mirror 'nuff said
*takes a breath* got all that?
There's alot more to add, but that's the medium answer.
