Reserving a high level slot is great general advice. Not necessarily with the expectation for healing, but with the expectation it's being reserved for one of those "oh crap" moments that can happen. Reserving potential for when it's really needed is specific to healers but if it's available a solid in combat heal can be important.
Agreed
Having it doesn't mean it's going to be used for in-combat healing, however. Out-of-combat healing options still tend to be more cost effective and using the same slots to mitigate damage generally saves more hit points in damage than the spell would have healed. Casting a healing spell in combat that could have been taken after the combat also nets a lost action for the caster so it only improves the combat actions if the healing gives the healed target more actions that would have been lost than the one the cleric does lose.
This is only part of the picture. Out of combat healing also competes against short rest hit die healing. So while, you may could heal more overall with an out of combat healing spell, that healing could likely have been accomplished without spending a single spell slot. If that's the case then using spell slots for out of combat healing is inefficient.
Now in situations where hit die healing is unavailable then it can make sense to spend a lower level slot on something like prayer of healing, but the general rule should be that out of combat healing spells are the exception, not the rule.
Let's face it. A 1st level spell the prevents attacks like Tasha's Hideous Laughter or Command or Entangle in a 1st level slot is going to prevent more damage from an Ogre than Cure Wounds in a 1st or 2nd level (until +4 bonus) slot even if it prevents only a single attack. Mass action denial is even more effective than healing using the same spell slots. Or go with defensive spells like Sactuary or Protection from Elements.
The downside is that action denial concentration spells must be cast at the start of the fight before knowing just how the fight is going. So let's look at the ogre example. Let's say his target is the 18 AC fighter with 20 hp. With his 6 attack he has a 55% chance to do 2d8+4 damage.
Tasha's has an amazingly high chance of success 70%
Entangle has an amazingly low chance of success 40%
I'm going to split the difference down the middle to estimate the average control spell will have 55% chance of success.
Now we have a lot of cases to consider:
Case 1: The depuff works every turn in the encounter leaving the ogre action less
Chance of occurrence: 17%
Chance ogre would have missed all attacks anyways 9%
Chance ogre would have hit with 1 attack 33%
Chance ogre would have hit with 2 attacks 41%
Chance ogre would have hit with 3 attacks 17%
1.66 healing spells used vs 1 control spell used
-----------------------------------------------------------
Case 2: The bebuff works every turn but the last leaving the ogre with a single action
This case results in an average of 1.66 heal spells being used to the 1 controllers spell
Chance of occurrence: =14%
Chance Ogre hits on the single attack 55%
Chance ogre would have missed all attacks anyways 9%
Chance ogre would have hit with 1 attack 33%
Chance ogre would have hit with 2 attacks 41%
Chance ogre would have hit with 3 attacks 17%
The debuffer in this case will use a healing spell when the ally is hit (this way I can compare apples to apples - number of spells used for the same result)
This would be 1.66 heal spells to the 1.55 debuffer spell
------------------------------------------------------------------
Case 3: The bebuff works only on the first turn leaving the ogre 2 actions
Chance of occurrence 25%
Chance Ogre hits with exactly 1 of his remaining attacks = 49.5%
Chance Ogre hits with both of his remaining attacks = 30.25%
Using the same logic of healing when the ogre hits.
1.66 healing spells used vs 2.1 control spells used
------------------------------------------------------------------
Case 4: The bebuff misses entirely
Chance of occurrence 45%
1.66 heal spells vs 2.6 control spells used
Weighting the averages
1.66 average heal spells used vs 2.08 average control spells used (for same effect).
(Not accounting for the extra cantrip attack the healer always gets on the first turn since no one needs healed then, or the extra cantrip attacks the controller gets when the enemy is debuffed and but would have otherwise hit - all in all I think those number of cantrip attacks for each pc will be pretty similar).
Amazingly, an average debuff spell at low levels isn't actually better than just healing your all back after being hit.
Given a choice in combat, I'm always going to go for what's effective, efficient, or necessary at the time and balancing that out with resource management over time. I find healing in combat something that may be required at times but it's a lower priority than preventing damage in the first place, which is where I see the best use for the actions and spell slots in combat.
So you are going to start healing in combat now instead of debuffing?