U vs R: It depends on the construction (e.g. door vs wall). For a glazed door or window (glazed subsurface), one can rely on the SimpleGlazingSystem option (product U, SHGC, VT). For an unglazed subsurface, a construction with a single composite (massless) layer may be adequate (product R, roughness, absorptances, etc.). Sometimes, all one has is a lab-tested U-factor from the manufacturer. But E+ needs more than that to describe a product's behaviour with respect to solar loads, long-wave radiative and convective gains/losses, etc. The links offer default values to consider for parameters other than U or R. The requirements are similar for multilayered constructions (e.g. of walls or roofs), although it's preferable to disregard massless materials in favour of standard opaque materials. This allows E+ to calculate the transient response of each construction.
Thermal bridging: It is generally expected that users describe a multilayered construction (e.g. of a wall) while considering the effects of repetitive (or minor) framing, e.g. studs, Z-bars ... that thermally bridge (or derate) insulating layer(s). The resulting thermal resistance of the construction is often labelled as a clear-field R-factor - clear of major thermal bridging (e.g. columns, corners, additional framing around a subsurface). We've been maintaining an OpenStudio measure that autodetects such major thermal bridging from OpenStudio geometry and auto-derates linked surfaces, before OpenStudio launches an E+ simulation (see initial UMH post, or the online Guide ... where we dive a bit deeper on what I describe above). One can also access the measure from BCL (look for "tbd").
I set aside any notion that thermal bridging is strongly linked to airtightness. Both concepts are treated separately in OpenStudio/E+.