Hi Timon,
I haven't done what you are describing, but having a 2x4 Solsylva in front of me prompts some thoughts - a 57 in gantry would be much more susceptible to racking, so I would make the gantry side plates longer and add some gussets between the beam and the plates, which would also require the x axis to be longer to keep the same working area as well as add weight. Dual steppers for the X axis might be in order. The Solsylva plans do provide some good info on modification and extension.
As for a plywood gantry, you really ought to see one of these machines in action to get a feel for the forces involved when the gantry starts flying around, and how much they want to bend. Of course, you can always slow it down to a crawl, but then anything complex to cut is going to take forever. It's an engineering problem, meaning you have to make hard compromises with cost, accuracy, speed. A key factor is acceleration, not just rapids, and increasing acceleration is where the table starts to bend and shake. This says you want to keep the weight down on moving parts as well.
Maybe, though, you don't really need the accuracy and a little bit wandering from a plywood gantry won't be noticed - it depends on the kind of things you are going to machine.
The danger is once you've built it, are you going to watch it thinking, I wish I'd have gone some other way? At a minimum, lay out a realistic cost for the plywood version, and a realistic cost for the aluminum one. Shop around for metal or junk dealers, especially any that might be local - they might have a I or C channel that would work.
How you value your time is up to you - some folks love building these things more than they like routing things - as soon as one is built, they're off on another. That's great - If you are more interested in building the table than doing CNC, then taking your time and trying different approaches are what its all about. I know I've learned a lot from this kind of builder right here in this forum, and I am indebted to them. (Thanks!)
However, If you see the table as means to an end, though, and want/need to start making CNC parts soon, realizing later that you have to re-do the Y axis might be a bit demoralizing as well as increasing the cost.
Probably more than you want to know about my machine at
www.liming.org/cnc
Good luck on your build, and upload some pictures!