First, what was good... Starting from the concept of animating a presentation around the classic "trains math problem", this was original. You took the basic and boring word problem and had fun arguing it. Setting aside working with a mouse for this (hat's off as you stayed "on visual" as much as possible) most of this was fairly clean and relatively smooth so it kept the storyline clear and more or less stayed on model.
Now, biggest improvements with the least additional investment... For most of the visual, you've hit a nail with the advent of the pad. Other than that, most of the visual arguments would be questions of "style" not exactly whether or not it's going to qualify as art. Breathe... relax... Now, there were points (and if you just watch it yourself, you'll see them too) where it feels a little rushed here or a little sluggish there. Some of that is (admittedly) my own machine and connection, but some of it was too consistent. The remedy for this is to go ahead and finish the last few frames and double check before splicing it all together. Almost every animator in existence is guilty of rush-work or skipping little stuff somewhere, even thinking that no one will notice, so this isn't to beat you up about it. Another tweak to get big improvement with low investment is the audio. Your audio isn't bad. At the same time, it can improve, so if you haven't yet, I'd suggest you check out "Audacity". It's referenced and used all over the place here, on NG, and it's free. If you take the time to read it up as you get it, there's also a wealth of how-to info' and with just a little practical toying around with it, you can get the sound track out of the tunnels and clean any annoying side-effects of cheap microphones (or none at all) when you're chipping away at syncing up the audio with the video...
Over all, (as the label for three star ratings says) Not bad... In fact, I like it. Oh... BTW, the answer is that the trains (one or both) will probably derail en route and they never actually meet. :oP