Crazy talk.
The "hot air rises" philosophy has literally zero significance with uniformly compressed air in an established laminar flow. The defitition of a laminar flow is that each molecule follows the one in front of it wherever the path of least resistance is, with resistance caused by a volumetric expansion/contraction, surface imperfections increasing adhesion, or a sharp bend in the plumbing.
To the OP: what makes you think the stock end tanks direct more airflow through the center while the bottom feeding end tanks would spread the flow evenly? Have you recorded any sort of data with your stock intercooler yet? There may be no reason to change unless the stocker is cracked or something...
Just for posterity I'll throw an idea out there that I've kicked around. When I look at the stock intercooler on my LML, the first issue that pops into my head isn't the fluid mechanical optomization of the end tanks. The first thing I think when I pop the hood is d*** that's alot of tubing to push a highly compressible fluid through! Swapping one core for a bigger or better constructed core still leaves you blowing air into one end of a very long crazy twisty straw and waiting to see results at the far end. Not to mention all the big names charging $1500+ to provide a moderately improved version of the stock layout. They deserve sand ground in all their holes for pretending to engineer an improvement.
Since my first concern would be decreasing the volume between the turbo and intake valves, I'd probably rotate the compressor on my turbo to discharge to the passenger side (not even sure it's possible on a stock gt 37) and route the discharge into an efficient water to air core, then straight back into a custom 3" intake feeding the manifold. With a water/air core it's easy to feed the cooling water counter current to the air charge to maximize thermal transfer and minimize air inlet temps. Get a radiator to cool the water and mount it strategically so your engine coolant radiator is getting more clean air than before. Maybe get crazy and add an ice box so your intake air charge is the same temp every time you go to the track regardless of ambient air temp. The benefits far outweigh the downfalls and the biggest improvement is that your total intake tract turbo to engine is cut to maybe 36" of total flow path. I'd be shocked if throttle response didn't improve a crap load.
I've never done it because my LML in stock form has the power to hit the speed limiter pulling everything I own up the biggest hill in tennessee, and the people currently charging $2000 plus putting your VIN on some EPA hit list just to tune decent throttle response into an LML deserve sand in all their holes too. All I'm saying is I've taken alot of fluid mechanics courses at a really prestigious school for a degree I'll never actually use, and my brain says if you're not optimizing the overall intake tract flow there's probably not a $1500 gain in another front mount intercooler.