If I got bucked off and hurt from a walk in a relatively safe environment, I’d personally be leery of riding him down in a scary one - but that’s because I’m old and out of shape : )
I suggest shelving the idea of liberty work for now, and for the purpose you describe. Liberty work is for horses and humans to build communication and strengthen a bond, not a free-for-all-mach1-orbit-lazy-man’s-longe-school – they can get hurt badly quickly, not to mention really do a number on the footing.
Because I no longer bounce easily, I teach my horses to have confidence in me. This way, when I do something stupid in the future, they can cover my sorry behind and forgive me. Here is my recipe for wimpy out of shape riders like me:
I would set myself up for success by doing a slow, steady progression of easy trust and confidence building exercises. First on the ground, then ridden.
I would prepare a list of about a dozen SIMPLE tasks Dobbin can do easily and successfully on the longe, and in hand. Having both types of tasks is key. Of course, the tasks should be on par with his skill level - easiest tasks that are still tasks, but very easy to do. For instance, trot a circle then stop. Or walk a half circle, let me approach you while you stand, do 1/4 turn on the forehand. Or trot to halt transition, etc. In hand, maybe pick up a foot by word command, or back on an arc 3 steps (straight back isn’t a good option as some horses can use that against you in the future), or two steps of shoulder in, etc. Very easy stuff.
In a safe area where Dobbin is happy, I would start longing with the goal to teach the first 4 things on the list. No fast rules, if #2 is problematic for some reason, just jump to the next easy task (thats why you have a big list). Have a significant reward for each time Dobbin does well - whatever suits the horse best, be it a treat, or praise or scritches.
Practice this to the point that Dobbin anticipates the task and also the reward. Let him figure out the game, know exactly what is coming up, what he has to do, and what he gets. The reward should be good enough that he’s eager to perform and very interested in his reward. The more consistent you are (standing in the same place, asking for the same task at the very same place on the circle, stopping and rewarding in the same place, etc.) the faster he will learn.
Once he is doing the tasks virtually on autopilot, take him to the indoor.
Do this on a day that you have the place to yourself - but don’t feel the need to minimize the spooks (ie, if there is snow sliding off the roof or squirrels in the rafters, so be it, might as well rip the whole band-aid off).
If the entrance to the indoor is in proximity to your happy longing place, that’s an extra bonus.
Walk Dobbin up to the entrance of the indoor. Stop and look into the indoor, and then immediately walk away. Do this twice. Then walk up, pause for a bit, then walk away. Then walk up, pause longer, then walk away. If he’s calm to the point of boredom, walk him up to the entrance and walk in without stopping or looking, walk only as far in as you can safely and smoothly turn around and walk right back out again. If he flips out, pretend it isn’t happening, do NOT console him. If he calms immediately down, do it again. If he’s upset, go to your happy longing place and do some tasks with rewards until he’s chill.
Do the walk in - walk out until he can handle going in and out without flipping his lid. As he gets calmer about it, walk further in. But the trick is to be smooth and don’t stop and don’t respond to antics, no cooing, no coddling, no admonishing - walk him in walk him out, but only go as far as he can handle. Bonus points to you if you can sense he’s going to get excited before he actually does and turn him around and walk him out just as he’s thinking about getting upset.
When you can make it halfway into the indoor, then up the ante by stopping and standing in the middle. Let him look around but don’t longe. If he starts to walk off, immediately start leading him out of the arena. Let him think it was your idea to leave.
When you can ask him to stand in the middle and he tries hard for you to be a good boy, then ask him to do one of his simple in-hand tricks and give him a monster reward. See how far you can go by asking for the task again and again and keep rewarding him. Depending on how many tasks he’s learned, keep asking and rewarding. Depending on how smoothly things flow, start longing and doing your longe tasks and rewards. If he gets scared on the longe, calm him down safely, bring him into to and ask for his in-hand task and reward.
The tasks he’s learned are his mental safe place. He has learned through repetition performing those tasks is being obedient and obedience is a happy place to be mentally. He has confidence in his ability to do them, and knows exactly what he’ll get in return. He in turn looks to you as a good leader because he is being obedient to you in return for protection from the boss mare (or gelding, I don’t know who you are!) - you are leading his brain from distress to happiness.
Once you are mostly on a good wave length and can longe and perform your tasks and both feel confident, then try riding. Again ask for only simple things from the saddle and reward big. Don’t get mad at him for being scared, and don’t baby him either, expect him to learn to have confidence in you.
This whole ordeal may seem glacially slow, and it is, but going slow in the beginning usually translates to achieving more faster later on.
Good luck!