You actually don't want to tie the xmbr to the tongue along the athwartship direction (as the existing plates do). Because of the constant bending moment at this location you really only want to join the members along the longitudinal axis, lengthwise on the tongue. (I made this mistake myself, fully welded my tongue tubes to the front xmbr and hope that I don't come to regret it later.)
If the existing tongue members are adequate, adding doublers won't really solve anything and will just add weight. Unless you plan on increasing net weight and want to increase your margin of safety, I wouldn't bother. And even then, I would rather see you add more of a truss like strap or rod (like Bob Henry used to reinforce the extended tongue on... I believe it was... his barn build), as I think that would accomplish the same thing with less weight added; and it would be easier to fabricate and fit. (Rather than notching a big piece of tubing to fit, make a couple of post standoffs located before and after the xmbr, then run a strap or rod from a location near the front of the tongue to the front post, to the rear post, then back up to a point further back on the tongue... like the main cable stays of an upside down suspension bridge. Sometimes you see this technique being used on the top of a P/U truck bed racks to help strengthen the cantilevered section over the cab. It's the same idea that all of the newer UT's use with lighter main rails strengthened by the separation of the side top rails... it's just an open truss rather than a big old honking piece of steel.) The bending strength is in the separation of the outermost fibers where the most tension occurs, not so much what happens in between them, and you already have the compression side covered by your existing members. (Okay, someone can argue that the tongue member will see dynamic loading in both directions, flip flopping from tension to compression, but it is all relative.)
The 'C' shaped channel vs. an [ (bracket symbol) shaped channel does add rigidity in bending (beam strength) but more importantly in this case, I think it adds better resistance to bowing/twisting. Standard 'S'-shapes (the ASME/ANSI/ASTM standard name for hot rolled [ shaped steel with tapered flanges) and simple formed sheet channels (as well as angle shapes) tend to fail by twisting and/or folding into the open side when bent. By forming the tongue members into a closed 'C' shape they have gained more strength while still being able to keep a lighter gauge material. In mass production shops they will often design with a bias toward fabricated sheet metal shapes (rather than hot rolled or tube) because they can better optimize the overall material weight, yield, and fabrication process. When I worked for Giant-Vac it was often more efficient to shear, punch and form (high speed machine operations) than it would be to saw, cope, and fit; especially in a bolted application. With a vertical break press and offset dies, it is relatively easy to form these closed leg channels, whereas hot rolling closed section tubing is not something most fab shops are capable of doing. By keeping operations in house and standardizing material usage, they can optimize their cost/lb on steel and stock fewer items (i.e. a pile of this gauge sheet, a pile of that gauge sheet, and not 100 different sections shapes and wall thickness' of tube). So again, I'm thinking your problem is the joint design and not the individual member cross sections.
One more thought. If you don't see any similar damage on the curb side, I'm wondering if perhaps the damage was not caused by a single overloading event; such as jacking the frame unevenly by the front street side corner; or touching ground, or an obstacle there while underway. A jackknife might also cause this kind of damage. (Sorry, I did not go back and refresh myself on the history of your trailer/build... though am sure I have followed along in time.)
I guess it all depends on how the weight of the cabin is transferred thru the xmbr into the tongue. One could argue that the tongue is pushing up on the underside of the bulkhead so the xmbr should bow center up, not center down (i.e, putting the lower flange into compression, not tension). The cracks starting at the bottom suggest otherwise.
Are there any significant scuff marks in the paint on the underside of the street side front corner? Probably not, the cracks do look more like fatigue failure.
Grind a small V into the underside crack first, then weld it, then grind it flush so that the "band-aid" plate lays flat. If there is any access at all (which it looks like there isn't) it would be nice to also weld the top side of this crack, at least at the ends of it (an option would be to grind a bigger v-notch with a root opening and place a small backer plate on top of the lower flange (slipped in under the bottom of the tongue... there appears to be a suitable gap there) so that when you weld from underneath it all becomes one).
If you (or your welder) have the ability, you could make the "band-aid" plates out of angle (or bent plate) with the front vertical leg cut down short to pass under the tongue. Even this short vertical lip would add back a bunch of bending strength (static moment of inertia, actually) to the xmbr... just don't bother welding it to the tongue, it will likely just crack there anyway... to be expected.
On the vertical weld, outside of the tongue, try to remove as much of the original weld, w/o scarfing into the tongue, before re-welding.
I see that I have "barfed" a lot of run on thoughts here. Hope it helps you more than it confuses. I'll check back in later tonight (to proof read again) and see if it all still makes sense.