Premise Behind Most D&D Modules

I've been gearing up for running G1: Steading of the Hill Giant Chief for about 13 years. My players are finally as close as I can get them. They established their base in the cave outside the Steading and are now using surveillance on the place to see what's going on and how they can approach this problem.

Here's the problem: They don't want to go in! They want to draw the giants out in the hope of taking them on in small groups and eliminating the steading by setting it aflame. Although this tactic is not unheard of, I could not stress enough how utterly wasteful such a tactic is. They think that eliminating the giants is more important than finding out who is the intelligence behind the attacks. How completely demoralizing for me. I was so psyched to play a good old fashioned dungeon crawl and they want to play siege warfare.

I suppose I may have led them on this path by allowing them all the gear and mercenary help they would require for free from the Duke. However, I envisioned them hiring henchmen, not soldiers. I kept asking them if they wanted to hire henchmen. They refused, instead taking up a 10 crossbowmen and 10 light hobilars with 2 sergeants. They thought they were approaching caves full of giants - they had no idea they would be finding a wooden fortress in the mountains. I allow the characters to make their own assumptions and plan their own attacks, only giving them warnings about how the system works in relation to their plans. If they attack the steading in the manner they are planning the party will be wiped from existence!

First off it is specifically stated that the steading will not burn. The timbers are so soaked with moisture that they will take enormous heat for extended periods to catch aflame. The druid's solution: wall of fire. We already house ruled the call lightning spell to be used once/round for 1 round/level with concentration maintained throughout the duration (still needs to be cast outdoors). I believe the druid is looking to cook the giants with lightning bolts as they emerge. With three gates, they can be surrounded and the dire wolves set upon them before they can get very many spells off. Even if the mage manages to get slow off in time, their deaths will simply be delayed, not prevented. If the entire steading is roused, the allies of the giants will likely participate in the attack - stone giants and a nasty cloud giant. These opponents are the deadliest. They are the most likely to make saving throws, the most likely to cause mortal damage, and the most likely to hit with boulders at range. The chief alone could eliminate all the henchmen with his ballista-crossbow. In the end it will end up being a rout at best and a total party wipe at worst. Either way my campaign will likely come to an end. I see no good that can come of this.

To step in and explain the situation to the party goes against my principles as a DM. Almost ALL modules written for D&D were DUNGEONS. They were meant to be areas to enter and loot, not eliminate at the front door and sift through ashes for surviving items. I'll give them kudos for coming up with a plan to take as few casualties as possible, but that's not the way the game is meant to be played. Wilderness is little more than a means of getting from a safe place TO the dungeon, encounters and all. Very few adventures are strictly wilderness based, and even then lairs are akin to dungeons so the concepts remain true. They'll have to accept this fact and run the scenario as it was meant to be run. I WILL have my fun, damn it!