I have never understood the grease argument. People seem to hold it up as some sort of ideal for creativity and playing outside the written rules, to the point that it has basically become a sort of shibboleth to sort who is a fun person and who isn't (like in this blog post), but most people I've seen know about the combo because they heard about it somewhere, not because they came up with this awesome original idea by themselves.
And, while the grease spell doesn't say it is flammable, it also makes no sense from a real-world application. Cooking grease at room temperature does not burn. the whole thing is more like the example of the flanking rule that is mentioned in the post. Everyone knows about the "setting fire to grease" thing, so obviously it's an actual rule.
And as for it's use as a shibboleth, I must admit that it works in my case. I am indeed, a no fun person. I like having rules to look up and will not allow much "creative" use of spells outside of what is written in them at my table. But, in my defence, I'd say that it is infact the other side, this strawman that I am now constructing, that is in the wrong. I feel like people at this point are basically treating spell slots as "plot tokens". "Hey, can I do this thing that normally I wouldn't be able to do? I'm willing to give up a 3rd level ~~plot token~~ spell slot for it." And there's nothing wrong with plot tokens, plenty of games use them to great effect. But spell slots aren't designed to be used like this in 5e, and it shows. It is (I think) one of the reasons why people think martials suck: They don't have any plot tokens to bargain with. If everyone had them it would be fine (Oh, hello, DnD 4e), but they don't.
No, I think spells should do what they say they do. Playing "outside your sheet" is fine and good, but you have to keep in mind what you are allowing when you do (and allowing creative ideas to be a great boon, but you can't repeat the same creative idea more than one is a good compromise).