Think about improving your pace in each discipline rather than your total time. Triathlons range in distance, and the distance you did is pretty atypical. It is one thing to compare times from olympic distance races that conform to the standard 1.5/40/10k distances, and another thing to compare total times from sprint triathlons where the swim and bike are wonky distances. Transition times are also a confounding variable, further reinforcing the point that it is best to compare pace in each discipline.
Swim: 400m in 8:52 works out to 2:13/100m.
Bike: 22k in 36:50 works out to roughly 22.3mph
Run: 24:03 5k works out to 7:45/mi
Relatively speaking, the swim is your worst discipline. People have VASTLY different definitions of what constitutes “Front of Pack,” and will badger my definition for this all day long, but the top guys here in NC are throwing down somewhere around 1:08/100m, bike pace between 25 and 26mph, and somewhere around 5:40 miles to the tune of a mid-17 5k. The exact numbers there don’t really matter because we’re only looking for a ballpark comparison. If you race someone that throws down those paces/speeds, the swim takes ~4:40, the bike takes roughly 31min and the run takes…say 17:45. On the swim you’re giving up a little over 4 min, on the bike you’re giving up just under 6 min, and on the run you’re giving up a touch over 6 min. Of course, the disciplines each take a different amount of time, so you’re ~100% slower than the FOP guy on the swim, ~10% slower on the bike, and ~25% slower on the run. Percentage-wise, you have the most room for improvement on the swim, 2nd most on the run, and 3rd most on the bike. You’re losing about the same amount of time on the bike and the run, because the bike is the longest leg of the race, and thus a smaller percentage improvement there will result in more actual time gained.
I’m not saying that my hypothetical FOP paces are anywhere near the limit of what humans are capable of, but they’re in the diminishing returns section of the improvement curve in the sense that each additional unit of training/effort input will result in less improvement per unit. Given that you are so far from this point in the swim, I would suspect that you will see the most improvement per unit training time input there. Accordingly, you would probably see the second most improvement per unit time on the run, and 3rd most on the bike. Improvement on the bike can also be purchased, unlike improvement in the water or on the run.
Transition is also an often-forgotten discipline where real time can be gained (or lost). In a non-wetsuit race that doesn’t take place in frigid weather, you should be able to spend less than 15 seconds at your rack in each transition. Cutting out extraneous actions, gear changes, and movement is the key to a fast transition. I am a firm believer that the fastest suit anyone can race in for a sprint or Olympic distance race is one like the Kiwami or PZ3–a suit that functions as a swim skin, but can also be worn during the remainder of the race. Having to stop in T1 to take off a swim skin, or put on a shirt/top of some sort will ALWAYS be slower than not having to do either of those things. Drying off hair/feet, putting on socks, and consuming gel/fluid are all extraneous actions that are not necessary and waste time when compared with someone who does not do those things. If you happen to do any of the above, I would recommend taking steps to change your habits such that you do not have to do any of those things in transition, because you are guaranteed to lose time doing those things when compared with someone who does not. In T2, stopping to tie your running shoe laces will always be slower than a cheap pair of speed laces.
Finally, some will disagree on this point, but it is my belief that rubber-banding one’s cycling shoes to the bike and properly executing a flying mount will save time over putting on one’s cycling shoes in transition, running in them, and executing a mount of any kind. The only way the rubber-banded shoes method is not faster is if you can not do it properly, thus the primary (and valid) argument against this method is that it is slower in actuality because most triathletes don’t practice it enough to actually save time using it. Thus the conclusion I’ve reached is that rubber-banded shoes (properly executed) is faster than putting them on in transition and running in them, while improper execution of the rubber-banded shoes method is frequently equal to or slower than the alternative. If you’re spending 2 min in transition and the transition doesn’t involve running 400m or more, then you should be able to cut some serious time off by eliminating waste.