Those are all questions/thoughts I’ve had myself. I too don’t have a large amount of capital to invest up front so I went with the option to do some work for a little period of time (a few days), then unbond and recoup my ETH to then be able to do work for a little longer period, rinse and repeat.
I set up my bot to only work jobs if the KP3R/ETH rate was below a certain level to prevent a big loss if the price decreased too much over the unbonding period. Seeing that it’s near its ATH right now, I would not suggest on paying gas for KP3R.
It was working well, but then the UniswapV1Oracle job was swapped out for Keep3rV1Oracle so I could no longer do work. The other jobs are still workable, but not nearly as often, so overall profit would be slashed. Seeing that the Keep3rV1Oracle job/MetaKeep3r is pretty much the only way to ensure you don’t run at a loss, keepers are kinda being pushed into the large capital investment. I’m not a fan of this move and proposed a change ([Proposal] Reduce minKeep on Keep3rV1Oracle to 0), but it appears nothing will happen (or really even be discussed).
So it seems this project is for the whales at this point and being a keeper with a few KP3R does not make sense. A few KP3R = speculative asset, a lot of KP3R = utility.