I wouldn't necessarily view this as the teleconnections are "wrong", I think it's important to understand what they physically mean, where they came from, and what they're really meant to represent in a larger sense. Teleconnections are merely representations of forcings (variance in mass transports/redistribution related to Rossby Wave behavior, external forcings, etc.) that ultimately drive them, and were made to simplify and describe the highly complex, non-linear, seemingly chaotic nature of the ocean-atmosphere into a few, easier to digest variables. To get a better sense of what's really going on here it's important to understand the physical mechanisms that can modulate them and understand that few (if any) of these teleconnection patterns are
truly independent from one another (although eigenvector (EOF) analysis may argue otherwise) and often possess significant overlap). For example, the North Atlantic Oscillation (NAO) mode is just a localized version of the Arctic Oscillation (AO) (thus the NAO is often (but not always) moderate-highly correlated to the AO and of the same sign) and the NAO just simply represents the meridional variation in the North Atlantic storm track, which is influenced by a plethora of phenomena (including but not limited to: Rossby Wave trains emanating from tropical convection, quasi-stationary planetary waves (w/ some seasonal modulation by ENSO), higher frequency Rossby wave packets and transients, and solar activity, just to name a few). Knowing what these phenomena are and what influence they could have on teleconnections like the NAO will provide you an appreciably deeper understanding of what's actually going on here and you'll probably see why I chose to disagree with your statement that the teleconnections were "wrong". I'm not saying they're "right" either, at least in the same sense as what you're trying to get at, but I'm trying to steer you in a direction that entices you (& others here) to take on a much different, analytical perspective of teleconnections if nothing else.
See Wallace and Gutzler (1980) for a more detailed, technical description and overview of teleconnections and how they were derived.
Although it's an old paper, it definitely helps give some perspective on where indices like the AO, NAO, PNA, etc. come from and what they physically mean/represent in a larger sense...
http://journals.ametsoc.org/doi/pdf/10.1175/1520-0493(1981)109<0784:TITGHF>2.0.CO;2