That if something is categorically imperative, that you have a duty to act from it regardless of convenience?
Actually, that's not entirely clear... but it's just a question of how we define "deontological" rather than any deep philosophical thing. Contemporary ethical discussion seem to love the term as a contrast with consequentialist theories or virtue theories. Moreover, Kant could not have used the term because it did not exist until after he was dead.
Yes, the CI is such that it obligates the moral agent to perform its duty and since it relates duty to morality, it is "deontological".
But then if we look carefully, doing your duty is not enough for Kant, the maxim of the action of your duty must be universalizable and in accordance with pure reason. There's a lot of hoops there, but to say it another way, if my duty were somehow to help an old lady across the street, I only do my duty for Kant when I help the old lady across the street from the motive of duty itself (rather than allowing some emotion or desire for praise to be the source of my duty).
This term is less helpful than the counterpart consequentialism, because different consequentialisms differ not primarily in the methods -- calculating the best consequence given some limitations but in the quantity that matters. Deontologies differ in the way that duty relates to morality at all (is it performance? is it rights-based? Are these duties we contract either through some social contract or as a condition of rational action in a world with multiple moral agents?) In many contexts, the differences in forms matter greatly.
(I have deleted the aggressive overgeneralization of Kant into psychoanalytic framing.)
It is more important that the imperative be categorical, than that it be tractable.
Otherwise we allow for horribly abusive directives that folks often actually act on like 'Every baptized Christian soul is worth preserving' which means 'Kill all the Natives, in case they might kill a Christian'. To the degree the clarity of the logical form is paramount over applicability, the actual absolute implementation of a moral rule might not be feasible.
Kant handles this with an analysis of absolute vs contingent duties. Absolute duties can always be obeyed. Contingent duties follow from my position as a being who has needs. Acting on them is to some degree subject to convenience, or we would all become neurotic messes.
I cannot stop for every homeless person. The society cannot keep everyone alive on a respirator who has been brain dead for years, or we would be unable to treat new patients. So (to my mind, rather softening the position) the categorical imperative is relative to my ability to imagine myself in equal need. If I were homeless and only one person gave to me every day or so, I would be OK, if angry. If I were uncommunicative for a year, I would not want the burden on my family to continue just in case I happen to be a 'Quinlan case', etc.