Slot-wise, was it worthy ? I mean it implies to sacrifice slots to do that.
Yes and no. Prior to getting your full complement of slots, I'm advocating burning one more slot for endred in four or five attacks at a slot cost of say five slots. At the same time I'm advocating pulling two slots each out of three toggles (in my SR example above). That's a net cost of burning one less slot.
Now, in terms of maximal slot configurations, you're theoretically capable of having six slots in each attack, which in the pre-Invention world is going to be filled with SOs or HOs (because the crystal titan enhancements would be dead by level 50 when you have all your slots). So let's look at some possibilities for a six-slotted attack with SOs.
There's six slots of damage. Pre-inventions and pre-ED you're looking at an attack that does (even SO numbers) 300% of base damage (+200%). But its also going to miss a lot: your base tohit is going to be 75%. To simplify the analysis I'm going to ignore things like tactics and intrinsic accuracy, just to keep the comparisons reasonable. In this scenario using this attack as often as possible you'd average doing 2.25x base damage factoring in misses (I'm also going to exclude the streakbreaker for now, also for simplicity sake). Let's just say this means this slotting pattern is worth 2.25.
The more common slotting was one acc and five damage. In this case, your chance to hit goes up from 75% to 75% * 1.33 = 99.75%, but that gets capped to the 95% tohit ceiling. Damage slotting drops from +200% to +167%. Total damage becomes .95 * 2.67 = 2.54x. That's better.
Now, some people went further and slotted one acc, four damage, one recharge. This gets a little tricky. Accuracy still hits the 95% ceiling. Damage drops to +233%. But slotting +33% recharge doesn't improve damage per second by 33% because of how recharge works. Recharge slotting improves the total recharge time of the attack but doesn't speed up activation. If you slot 0.33 recharge, you will reduce an attack with base 4 seconds of recharge to 3 seconds. But if the activation time is one second, the power goes from taking 5 seconds to cycle to 4 seconds to cycle. That's only a 20% improvement. If we guestimate that the average attack has 1.5 seconds of cast time and 6 seconds of recharge, then one recharge improves cycle time from 7.5 seconds to 6 seconds, which is an improvement of about 25%. So the net damage per second of this attack, judged alone, is about .95 * 2.33 x 1.25 = 2.77x. That's even better.
But man, it is costly to run. If you are cycling this attack as fast as it recharges, then (ignoring Arcanatime lag for simplicity sake here) you'd be burning 6.864 endurance every 6 seconds (an attack with recharge 6s would have damage 1.32 and endurance cost 6.864). This one attack alone would burn 1.144 eps. You'd be hard pressed to sustain four of these, which would be the minimum necessary to reach a full attack chain (it would probably require slightly more than that). So let's see what happens if we slot one endred.
One acc, three damage, one endred, one recharge. The damage multiplier becomes 0.95 x 2.0 x 1.25 = 2.38. That's less than 1acc/4dmg/1rech but it is actually better than 6-slot damage and 94% of the damage of the 1acc5dmg slotting pattern. That's not bad. More importantly, endurance burn drops from 1.144 eps to 0.86 eps. That's even lower than the burn rate of the 1acc5dmg slotting pattern (0.915 eps).
So: if you compare 1acc/3dmg/1rech/1end vs 1acc5dmg, you end up with 94% of the dps at 94% of the endurance cost. That's actually pretty good and it is more sustainable. It only looks bad compared to 1acc4dmg1rech which deals more damage but is totally unsustainable.
And there's one more slotting pattern to consider. 1acc/4dmg/1end. At the time period we're talking about, perma-hasten was possible and not uncommon. Under perma-hasten you had a permanent 70% recharge buff. That made it practical to consider not slotting for recharge and only slotting for endred. When you account for the 70% recharge buff you end up (using the same assumptions above) with 5.03 cycle time, 1.49x cycle rate, 0.95 x 2.33 x 1.49 = 3.3x effective damage and 1.03eps. Compare that to 1acc5dmg which would be 0.95 x 2.67 x 1.49 = 3.78x damage and 1.37 eps. That's 87% of the damage at 75% of the endurance cost. That's a pretty good deal.
Given the "standard" slotting at the time of 1acc5dmg as the benchmark, 1acc3dmg1rech1end performs comparably outside of hasten. 1acc4dmg1end performs better than 1acc5dmg when under hasten in terms of sustainable bang for buck. So when slotted as part of a comprehensive build plan to manage endurance and damage output, endurance reduction slotting paid for itself. It wasn't a sacrifice in the sense it actually opened the door to dealing more damage over time in a sustained fashion relative to not slotting it. You could temporarily deal more damage for a short time without it, but you could deal more damage overall with it.