But the question here is not whether giving to charity beats acting romantic to one's partner (Gleb and his wife are obviously being romantic to each other; indeed, they're also choosing to enjoy an experience which will likely make them happier in the long term - dining in a nice cozy restaurant), but whether it's better than buying expensive stuff for themselves.
I don't think that's the question. You aren't constrained to the options of an "expensive" gift (carrying the connotation of low emotional resonance) or a charitable donation. You can also spend that $50 on another nice experience with your loved one, or you can buy a cool accessory that goes with their sense of style, or you can buy a beautifully bound journal and fill it with thoughts you have about them over the course of months, etc. You can do a lot of things. I'd guess that around 100% of people I know and >99.5% of genpop would find one or all of those options more romantic than a charitable donation, and that it would make them happier. I have no churlish objection to this particular couple finding a donation to be the most meaningful possible gift. But the overwhelming majority of people won't, so presenting this as the option that'll make them happiest is likely to fall flat.
And the evidence seems to be that getting a costly material gift raises the giver's status in your mind, but doesn't really make you happier.
I haven't seen this evidence--can you link me? Nothing in the post, or linked in anything linked in the post, seems to show that. Does this evidence apply to romantic partners, and does it include "personalized" gifts? (I.e., those the giver put a lot of thought into, and which were selected specifically for a partner they know very well.) I would be quite surprised if that were true. I would be completely unsurprised if it was true of arbitrary material gifts, but I think the takeaway there is "don't buy your wife a washing machine for your first anniversary" rather than "don't buy your wife a gift."
The effect of gift-giving on the giver isn't to be neglected, either. Doing something nice for a loved one makes most people feel particularly good about themselves. There's also some indication that such actions make most people feel more connected with and devoted to the recipient. (Essentially an extended form of the Ben Franklin effect,)
Harsh, but this does have two HuffPo-like traits: first, he uses his opening line to make a point that's grossly misleading, and repackages his generic pitch for EA as something relevant to an upcoming holiday. "Hey, you know what's the most romantic thing to do? Turns out that it's the same thing we recommend doing all the time. What a coincidence!"
Second, his factoids about the psychology of generosity are as misleading as HuffPo-tier science reporting. Generally speaking, the psych/neuropsych studies I've read don't really support the conclusions that EAs seem to want them to, including those studies that they cite as evidence. Specifically speaking, in this case, the studies don't seem to indicate that charitable giving is special, broadly or vis-a-vis the activity that this post is contrasting them with. I.e., neither of the articles provide evidence that giving to charity has a particular advantage in making people feel good over other forms of generous behavior, including the conventional Valentine's Day one of giving something nice and romantic to someone you love. Indeed, most of the research I've seen on the subject indicates that a wide range of actions taken on behalf of others produce neurological rewards.
I'd find it very strange if actions toward other people you didn't know produced greater psychological rewards than those you knew and loved, and I've yet to see any evidence that it's true. Anecdotally, it seems vastly more likely that the opposite is true: that if you're trying to maximize your own happiness, being generous to the people you love is the best way to push this psychological button.
Regarding your first point, which do you suppose is more likely: that love is a bad idea, or that having a very reductionist viewpoint is a bad idea?
Regarding the second, a lot of things are like "the brain chemicals involved in love." (The article only discusses low serotonin levels.) This doesn't provide a basis for thinking love is a bad thing.
Regarding permanence, "is lifelong commitment to a single person a good idea" is a different question from "is love a good idea?" Since you've asked, though, I think I disagree with the mechanics you describe. The benefit of a lifelong love isn't strictly limited to the loved one as an object, or the traits inhering to the loved one, but the interaction between you. This grows over time, while other interactions that you aren't having are likewise not growing.
If the argument being put forward is that it's not good to give terrible romantic gifts (i.e., those that make neither the giver nor the receiver happy), and that, as such, any nonnegative alternative activities--including charitable gifts-- might be better, I find that very difficult to disagree with. Personally, I think that the correct response to that situation is to get better at giving gifts, though.