Still havent seen one person tell me why its a bad idea other than "shit happens" and "people want to play engineer"
"People want to play engineer" is a great reason. This is, after all, a game that people play because they want to. I fully agree with you that 5 engis was crazy, and meant TPG didn't get the
expected result out of his test, but it's still a valid test result: it shows that his map, and/or gamemode, is vulnerable to engi camping.
That's valuable to know, since this is one of the key reasons 2fort fails (although not the only reason it fails so spectacularly for comp).
And it's fixable: some gamemodes don't suffer this problem at all; the layout can be designed to make such strategies ineffective, so an engi-camping team would be prone to losing; or as a last resort, key areas can be func_nobuild'd.
If you are designing your map solely aiming for 6v6 comp play and don't care about 12v12 pub play, then you are wasting your time testing it in 12v12 pub. But if you care about pub play, then learn what the tests teach you. The key to playtesting is to take the results you get and analyze them, not to arbitrarily change the rules so that result doesn't happen.