Just a quick look off that image your secondary mirror isn't square to the focuser.
Also laser collimators are generally not collimated themself and the red dot is also a sausage makes it a bit hard. So first of all pull it apart and put a mask over the end of the laser so you get a tiny pin prick for a laser dot, then collimate the laser itself by rolling it round and round untill the dot doesn't move over say 6m distance from the laser.
Next get a ruler lay it across the spider vanes and make sure that the spider vanes are straight.
Open up the primary mirror cell and replace any springs with far bigger and stronger springs (this is the most probable source of your change in collimation).
Next. open up your focuser eg have it as a 2" straight though, and look at your secondary. it should present a circle that is centered within the focuser up down left right.
Now is when you put your laser in and do its job. get the dot to land on the middle of the primary using the secondary to get it there. then use the primary to return the red dot back to the collimator. This makes sure that the Secondary, Primary and Focuser axis are in line.
Hopefully that gets you in the ball park.
For better collimation you need different tools and my preference is no lasers
Auto collimator, Cheshire and Sight tube will give you virtually perfect collimation No guessing and no star testing, as star testing newts is a very tricky and hard task with every mirror floating!
