There's no comparable real world situation where selling a city that is about to be taken by an enemy army could possibly happen, it's just absurd when you try to relate it to real human civilization.
It's just absurd when one uses games a model for real life and then tries to compare the two...which carries the same logical weight as using real life as a model for deciding the credibility of in-game tactics. For example:
People live 1000's of years, or see cities (with no military at all!) firing at siege-level ranges, or plan out exact technological paths of unforeseen breakthroughs that will develop their civ. Real life isn't like these things; therefore real life is ridiculous!
If players don't like that games make concessions for gameplay purposes instead of conforming to reality 100%, the game doesn't have much chance of being fun. If players DO accept those necessary concessions, than decisions on what is "cheap" and what is "good play" in games like civ V are COMPLETELY AND UTTERLY ARBITRARY
.
Avoiding tactics like this for a challenge or personal preference is valid...the basis of it being "cheap" and "not realistic" is not valid.
Firaxis put this beta out as it is...play by the rules
!
Of course, there ARE gameplay arguments to to levy against this tactic (and they are the only remotely valid ones), of course. Probably the biggest issue is the allowance of military collusion without actual war; in MP this would be seen as an obvious alliance against the attacker, and nobody would begrudge the attacker from swiftly pounding the gifted city to oblivion. Since this IS a de facto alliance, the ability for the alliance to forcibly move opposing units without killing them or using actual military power is concerning (city gift of this nature should be treated as an act of WAR by the guy accepting, and treated accordingly). If this were able to draw the buyer into the war as an aggressor (IE taking a war bribe) and it didn't expel units, it would be balanced...
However banning tactics in a beta release parading as a finished title is inherently arbitrary. Look at the June patch and the #things changed...any argument of "this will probably be changed" falls doubly invalid by firaxian patch history (dating back WAY earlier than civ V; there is still lots of unfinished garbage in IV for example)...you can't make a legit claim that "this will likely be patched" OR "this isn't patched so it's intended" because the post-release support of unfinished features has been SO HORRIBLE that neither argument is reliable (quick civ IV examples: overflow, apostolic palace, the entirety of the vassal state mechanic, how AI calls UN resolutions, game controls).
In other words, this tactics "feels" wrong because it takes advantage of an incomplete AI in an incomplete game. Where do you stop though? Do you avoid fleecing the AI for gobs of gold? Do you avoid baiting it into declaring on you and then killing units nonstop to farm xp? Do you avoid RA spam that they can't match? Great people? Etc? None of these things, based on Firaxis patch history, are any more or less valid than city sale (this tactic ALSO existed in civ IV, mind you...................was it ever fixed? Take a WILD guess
)