Jed and CANPRO are correct. You usually cannot decrease the development length where the bars have to be fully developed. What you describe has long been done with headed anchors for shear applications, but I would not spec this for standard rebar, especially in tension, without some test data.
First issue is that, save for some friction, bar develops based on the deformations, not surface bond interaction. This may be assumed to account for 30%, based on the factor for epoxy-coated bar which has no surface bond with concrete. Making a hole for a grouted embedment will probably result in a relatively smooth hole.
Second, a high strength adhesive or grout could effectively increase the diameter (and thus surface area) which would be useful to decrease development length. For this to be useful, the interior of the hole would need to be rough, with an amplitude similar to that of the bar. That said, it has been found that simply increasing the height of deformations on bars does NOT effectively decrease development length.
Third, getting nearly 100% embedment with a grout will be difficult, if not impossible. Doing this with an adhesive is similarly difficult, but the relatively higher strength and narrowness of gap between the bar and the surrounding concrete mean that small voids have a lesser effect on strength of the bond.
Finally, it is not only the development of the bars you intend to embed, but also how the other bars inside the concrete interact with the concrete. It is insufficient to connect the new bars to the concrete if the concrete cannot get the forces where they need to be. ACI 318-11 App D has guidance on dealing with this.