gistfile1.clj

; update-vec merges a vector based on the :id's of its contents

(update-vec [{:id 1 :data "1"} {:id 2 :data "2"}] {:id 3 :data "3"})
;=> [[{:id 1 :data "1"} {:id 2 :data "2"} {:id 3 :data "3"}]]

(update-vec [{:id 1 :data "1"} {:id 2 :data "2"}] {:id 2 :data "tonystark"})
;=> [[{:id 1 :data "1"} {:id 2 :data "tonystark"}]]

How about below?

(require '[clojure.test :refer (with-test is run-tests)])

(with-test
  (defn update-vec [som m]
    (if-let [i (first (keep-indexed #(if (= (:id %2) (:id m)) %1) som))]
      (assoc som i m)
      (conj som m)))

  (is (= (update-vec [{:id 1 :data "1"} {:id 2 :data "2"}] {:id 3 :data "3"})
         [{:id 1 :data "1"} {:id 2 :data "2"} {:id 3 :data "3"}]))
  (is (= (update-vec [{:id 1 :data "1"} {:id 2 :data "2"}] {:id 2 :data "tonystark"})
         [{:id 1 :data "1"} {:id 2 :data "tonystark"}])))

HTH