By what measure is selling luxes for gold an "exploit"? I can't recall ever seeing anyone make that claim. Some people consider selling a lux before DoW or letting a lux get pillaged on purpose an exploit (something which I largely agree with) but just any old trade?
"Exploit" may not be the best phrase, since it doesn't do anything to substantially disadvantage AI rivals, and indeed strategically straight lux-for-gold sales may not be particularly advisable when maximising golden ages or supporting 4 more pop for the 30 turns of the trade can lead to greater returns. But it's lazy play since it takes no actual work - you can do it with anyone you aren't actively hostile with, and it allows you to churn out archers or spears (or cost towards settlers) without needing to put particular effort into developing your early economy.
You get at least as much science benefit from getting 2nd and 3rd cities up asap as building a library thanks to the population.
Not immediately - if, like me, you play to maximise early growth in the capital, that new library is an immediate +4 or 5 beakers per turn. Building settlers takes almost as long per settler as building a library before you unlock the appropriate policy if going Liberty (which I no longer find a very useful tree to take in most circumstances), and if you do build rather than buy them you're delaying pop growth. And by the time you've settled two new cities and grown them, your capital will be churning out still more bpt. You'll equalise eventually, but there's no substitute for an early library for getting ahead early (unless you're Maya, or at a pinch Babylonian or Korean).
Mining ain't just for building mines
No, but if you go Bronze Working (or Masonry, come to that) immediately after Mining, that's still longer to wait before you get your Library. If you don't go for it immediately, you don't need to research Mining that early. Attila and Darius aside, it's not a tech route I'd choose.
I get rushed in games before turn 70 far more often than not.
I rarely get rushed at all - the fact that I've been rushed in my last three games came as an unwelcome surprise, and even then not one of them was before turn 70 (on Immortal). Recently I've often been the one launching the rush in collusion with a friendly AI, against an unsuspecting AI victim.
I will grant that simply looking at hammers in < hammers out is not, by itself, a complete measure of value on a wonder, and perhaps I could have phrased my original post on the subject better. But it is entirely reasonable to say that if the number of hammers going in is far more than those coming out then the wonder leaves something to be desired, and IME that is the case with SofL.
I actually agree that it should have gone before it did, and I don't often build it myself, but there are plenty of things that enhance specialist potential that make maximising specialists a good strategy (and especially if you're Korea). There seems to be a mindset in Civ V that specialists are for Great People first and production second (and granted this is largely true of Workshop path specialists, since if you have a production city you're rarely going to exceed the output you could get working a tile with the mediocre hammers from an engineer). But most specialists produce output you can't get from tiles (and by the time you've got several market chain buildings, you're likely to be running out of gold farms that match the same tile output), and larger cities in tall empires will simply run out of tiles to work anyway.