I can tell you for free that 'designed with the idea of humans playing vs humans' is a false assumption. The rule for bundled scenarios is that they can be played in all modes although exceptions to that rule were made for CMCW in relation to the Soviet Tactical Doctrine primer scenarios and a couple of others in that title. None of those exceptions include H2H only. There was no direction given to any of the scenario designers to optimise their scenarios for H2H for any title I've been involved with (CMSF-1 &2, CMRT, CMFB, CMFR, CMCW).
The challenge of following the rule, particularly when making a scenario based on a real action, is that the scenario has to be winnable by both sides in all three modes. The get out clause is the one-liner in the 'Load New Game' screen where you can say 'best played as (insert side).' Designing H2H is even more variable:
How do you know how skilled every single player is who buys the title?
Which of those players is going to take Blue/Allied in your scenario?
Which of those players is going to take Red/Axis in your scenario?
Is one of those players going to play to their skill level or just have a bad day?
Is one of those players going to have a good day and play above their skill level?
Steve from Battlefront posted some time back that the data/feedback he has indicates that most people play the title in Human vs AI mode. This would likely explain why the rule of playable in all three modes was introduced. It also reinforces the point that your assumption is not well-founded.
Nonetheless, designing for all three possible combinations is achievable but it generally involves employing most of the victory point combinations, asymmetric objectives and time limits. Achievable of course does not necessarily mean that your scenario is awesome in all three play modes. It should be in one of the three and if you can do it in two, then so much the better. If you can do it in all three then your name is @George MC
Linking the above to your point that designers 'boost' defenders with 'tons of points' - that is correct in many instances. I 'boost' one side or the other or both sides with victory points in order to achieve the effect I intend. As an example, to avoid a turn one cease fire resulting in a victory for a defending force that typically occupies all of the high victory point objectives that the attacking player needs to capture I will 'boost' the attacker by giving that side the equivalent number of victory points for friendly casualties at a threshold that will only be achieved by the defender's actions half way through the battle (ballpark figure for illustration would be 20% casualties). The turn one ceasefire would; therefore, result in a draw. The intent here is to make both sides commit to the scenario and play it through. If both sides commit then the attacker will not get those victory points and is not intended to. You confirm these thresholds by testing and adjusting as necessary if that 20% (or whatever) threshold is achieved too early. Some of those 'boosts'; therefore, are never intended to take effect if the scenario is played with good intent by both sides through to its time limit or to a point where one player or the other genuinely elects to cease fire.