That is an interesting question.

I'm going to suggest that the lengths would not need to be increased due to the epoxy coating. My logic for that is that if you work from the tip backwards, the bar is still extending past the inflection point, and begins developing at it's tip. Whether it is epoxy coated or not, the bar will be fully developed well before you get to the critical section. The only exception to that would be if you are looking at very short spans, where 0.33ln or 0.2ln is less than the development length of the bars.

If the development length of the epoxy coated bars is larger, then they should increase by at least the increased development length.

Those are minimum lengths. There are a lot of situations where the length required is larger, eg, varying span lengths and pattern live load.

I've always found that diagram confusing. When you read the blurb below:

1) The red bit presents it as a minimum, somewhat implying that extensions should be based on analysis with the minimums just being the "floor" so to speak. Nothing explicitly says that detailing the rebar to the minimums constitutes detailing it adequately. I do realize that a great many firms do assume that extensions are adequate when the minimum extensions are provided however.

2) The green bit, in explicitly saying that extensions should be based on analysis for lateral loads, kind of implies that the the use of the table bar extensions is somehow adequate for gravity loads. Otherwise, why only mention lateral loads as needing to be detailed per analysis?

3) My understanding of the tabulated extensions is that they were developed in conjunction with the direct design method which has specific limits on geometry, loads, etc. Rationally, I'd only expect the tabulated extension to constitute adequate designs when the slab properties fall within the bounds of those limits. To my knowledge, there is nothing explicitly connecting the table to the direct design method in the current ACI code.

4) Excepting the above, and assuming a design for which the bar extensions would have been adequate as detailed in the table, I agree with ryaneng that introducing the epoxy should not much things up for top bars so long as you still have proper development by the time that you get to the critical section. In addition to development at the critical section, one also wants the partially developed tensile capacity of the bar to match the tensile demand at all locations in the slab. For uniform loads this should be self satisfying given the assumption of linear partial development in the bars as shown below.

i dont have ACI in front of me, but one thing to mention about the table is that it doesn't account for bar sizes or any other factors that impact development length. that implies to me that the table is more of a good detailing practice, to allow for a smooth transition between positive and negative flexure, based on spans lengths and typical loads. At the end of the day the development length still needs to be checked. The is dependent on reinforcement spacing, bar diameter, epoxy coatings, etc., The required development length for bar curtailment could potentially exceed what's shown in the table.

Having said all of that, there's a good chance for most typical conditions the table is more than enough, hence it's use in typical details.