The two quotes you use as proof speak about two different types of architects. Linthicum is correct in the first quote, EA needs to provide value, but value comes from the practitioner, not the concept. I have seen architects that do provide value by consolidating applications, correctly guiding business to select tools that deliver a strong ROI, and guide an organization to a position of innovation.

The second quote is specifically about software architects, a different type of "architecture". EA is a master planning discipline not only focused on one area like building software. Building an enterprise is like building a concert hall or factory and for that one needs a person planning the whole structure. Software architecture is like designing the HVAC system for the concert hall. The person in charge of that does not need to know about the geology of the site, lighting quality of the windows, but only about part of the needs of the building's users . . . but the EA needs to know about all aspects and help design all aspects of the enterprise (building).

EA is valuable if you put value into it and allow the EAs to specify most of the business elements like applications, processes, and infrastructure, etc.

Odds on the same people that complain about Enterprise Architects are the same ones that complain if there isn't a plan.

The challege with EA as role, is that cutting it out, does not cause immediate pain.  It's a pain prevention exercise.  Bad enterprise architecture is like bad eating habits, eventually the results will show and it will be too late to correct without a lot of pain. 

Many EAs have the knowledge and information and make good recommendations, but they lack the salesmanship for convincing leadership to follow their advice.   Advice that isn't followed is absolutely worthless.

Effective, enterprise architecture is 70% sales and marketing, and 30% strategy planning, process, application, information, infrastructure, standards, governance, integration and so on.

The best and most effective architects are the ones that understand business best practice, have street credibility with operations, are directly involved with project delivery, and can engage and sell at "C" level. Of course, it is all about politics. That's why they get paid well. They stick their neck out above individual agendas and business silos to champion what is best interest for the enterprise. Occasionaly, their head detaches from their neck.

Typically, these people come from 10+ years of consulting background with strong vertical industry experience and software vendor exposure. They operate as an internal consulting unit with the enterprise, earning their living by charging for their services. The question as to whether they sit inside or outside the enterprise payroll is not-important.  However, for large corporates with significant transformations underway, I recommend an in-house approach.

For those with an interest in the future of enteprise architecture, may be worthwhile reading "Enterprise Architecture as  Strategy" by Ross, Weill, & Robertson.

I am a chief architect and certainly question EA.  We act like we invented all this.  Strategic alignment, effective models, corporate standards, reuseability and convergence all existed before EA.  All we have done is repackage them and sell them back to our clients, again.  With a 70% failure rate its safe to assume those people who failed in the solutions space before are now failing under the guise of EA.  Even people who boast EA success are not transforming their businesses the way we suggest.  EA is a principle that we have translated into a program.  We also say we are the transformation agents but have not changed in 10 years.  I do oppose the notion that we are anti SOA......we take credit for it along with anything else we can loosely connect to EA.