In the 70's I spent many happy hours devising laser-beam steering mechanisms, electronics, software, etc. for light shows and other purposes, with laser-beam power levels of 1 to 20 watts. Back in the days when laser shows were completely unregulated, relatively basic and mostly hand-made technology cribbed together for the occasion was up to the task of writing legible text on a wall (or cloud).
This was also before the advent of high-power semiconductor lasers, so most of the light sources were great hulking water cooled boxes with 10 or 20 KW coursing thru them to make the beam, and water cooling to keep the temps down at the prevailing power efficiencies of 0.007 percent or so. They weren't portable in any sense.
These days, permits and inspections and such, along with thick rule books, impose tight controls on aerial discharge of light beams... with procedures specifically intended to prevent eye damage in the event of casual unintended contact.
Laser diodes have made the optical power source easier to carry and afford, but beam aiming is still very challenging at any long distance if done from anything but a superstable platform. Successfully putting a single beam through a small window aperture on a rapidly-moving target, in real time, and on short notice is not really much easier than it would be with a projectile weapon. Doing that continuously or even repeatedly for one or two seconds seems very nearly impossible to me, using manual or simple mechanical and electronic means for aiming.
The one simple technique that would "appear" to work for such a time duration is to scan the target vicinity quite rapidly with the beam (perhaps spread out to make it cover a wider surface) moving in a pattern that would tend to periodically sweep all points on a larger surface. This process of "waving it around" greatly dilutes the cumulative beam power on any surface, however, be it eyeball or skin.
Because the mechanism of damage from laser light is tissue heating at the microscopic scale, the cumulative intensity on a surface in a given period of time is the metric of danger for permanent damage. I don't have the tables handy, but my intuitive guess is that
100 mw at the source for example, would not be able to cause much ocular tissue damage after a few hundred meters of normal surface atmosphere, however well it might be aimed and focussed at the source.
With even 5 mw of distant visible beam, one certainly is very aware of being targeted. The brightness of frequent flashes can be subjectively quite spectacular, and certainly would be a worrisome distraction during night flight.
It might be some consolation to know that truly damaging laser light works so rapidly that one tends to not actually even see it.
More sophisticated methods, which I will not detail here, can be applied to provide near continuous tracking with a laser beam of a spot on distant moving surface, but the cost and complexity to do any of these in a field-able way would take them far out of the kitchen-table category, and the result even then would probably not be readily capable of reliably "blinding" the unhappy recipient after running through the atmosphere, window material, etc. Anyone with the motivation, skill and budget to produce that type of device would likely be able to devise much more dangerous effects with equivalent effort.
So, my guess would be that the actors in many of the inflight illumination incidents are amateurs...trying to find something interesting to point at with their expensive toys.