Critique is helpful. We're not telling you to give up and go away, we're saying that there are clear design considerations that you're either ignoring deliberately or are unaware of - saying "I am increasing the stiffness of the frame and gantry" without specifying how just leaves the whole thng open to misinterpretation - if you want valid design input, you need to offer more information than you have thus far.
For starters, you don't have a gantry - that's an overarching frame that's supported at both ends. You've got a cantilevered spindle instead, which is a much weaker system of attachment when it comes to dealing with the kind of dynamic loading that a mill or router has to deal with. Ordinarily, this is resolved by having as short a cantilever as is possible, which is why the spindle is never normally mounted away from the z axis like that.
Frame stiffness is also only part of the issue - how are you looking at vibration damping? Again, adding weigt is a great way of eliminating it as a problem, but the kind of airspace design you have at the moment has very few avenues available to it to be able to resolve that.
When I said to stop spending money right now, I didn't say to quit, I said to continue researching the project - I'd really like to see you succeed, but I genuinely do not believe that you're going to get very far ith this current design, because it's just not going to be stiff enough to have any degree of consistency or repeatability in the cuts that it makes. This is too close to a 3d printer design, and the requirements are very, very different. 3D printers don't have any kind of serious load attached to the printing head - they're depositing material in what's effectiely a zero-resistance environment (unless something's gone very wrong), whereas a mill or router has got a continuous dynamic load applied to it, because it's being forced against material that it's expected to cut away.
The pdf article you cited pertains specifically to toolbit deflection, and assumes a certain degree of rigidity in the CNC machine frame that yours won't have, due to having that massive arm hanging off of it. If the spindle were brought right back in to the Z axis, then it might be more relevant. As for the thread link, 20klbs probably isn't a bad figure to aim for, but you really do need to run an FEA test on your current design based on that. I absolutely guarantee that the spindle arm will need a thorough redesign.