First... above all... in today's way of thinking about QFT,  it should be impossible to make
your idea about observing the universe work... or impossible to find ANY OTHER WAY
to image the future, in effect. It falls out directly from the mathematics as they use it.
Given a huge telescope and temperature distribution and all... the predicted outcome
can be found by coding up the initial density function, evolving that forwards in time,
and applying measurement operators WHICH DO NOT ADMIT any dependence
on future time events. That is hard-wired into the measurement operators.

And... more precisely..

The events you are proposing to use, to image the future, are events in which a photon is connected
in the present to this telescope, and in the future to another region of the universe. But in today's view,
probabilities of photon emission into "free space" (which includes billion-light-year pathways if
it includes anything!) can be calculated entirely from the local conditions, using well-established
calculation rules; your predictions clearly violate the predictions of extremely well-established
calculation rules, rooted in quantum measurement (which really includes setup as well as
measurement in the usual sense -- at least they appreciate the crudest aspects of THAT symmetry).

Thus we need to propose a modification to those usual calculating procedures, in order to prove
that we are not either ignorant or crazy.

There may be some physicists who do not know enough about today's mainstream to see this immediately.
But we can't ignore that mainstream.

Now that I read your email... I am stunned by a possibility that I never even considered. As I read through Savitt's book,
I was struck by the way in which you really understood -- your chapter really does shine extremely clear
and correct light on the issues it discusses -- while the other folks writing or cited (even Leggett to my surprise!)
fell into such obvious and deep logical traps. I tacitly assumed people would immediately see the contradiction
between what you were saying and the version of Aharonov formalism discussed by Unruh and others.

But maybe not.

And in fairness... there are some subtleties here that we all need to be more careful with, including me.
It is sometimes hard to say what is "the theory" and what is (1) the usual use of it; (2) the intuition behind it.
Quantum theory generally jumbles these up anyway, because it is not really an axiomatic structure.

And so... It is standard practice to compute "matrix elements from state i to state f" (leading directly
to predicted transition probabilities) as <f|exp(iHt)|i>, where allowed final states f are defined
as "eigenfunctions of the larger system." But most of the backwards causal effects I have
been thinking about (in laboratory contexts, i.e. on earth)... can actually be fit INTO that sort
of standard framework... even though they violate the causality assumptions!

I have not yet studied hard how Hegerfeldt "explained" the Fermi paradox, "and showed there is no
real FTL or backwards time effect there" (as conventionally stated). But I suspect there is a strong
formal parallel to the VCSEL experiment, the charged-pion-exchange nuclear reactions, and to experiments
Yanhua Shih has done!

In effect... if allowable states f are defined as eigenfunctions OF THE ENTIRE UNIVERSE, computed by
accounting for absorbers and potentials far away from the initial emitter... perhaps even at astronomical distances...
the <f|exp(iHt)|i> calculations might indeed allow ... FTL effects, at least. But then again, if the
proposed sinks are moving objects, even the <f|exp(iHt)|i> framework is not quite enough.
(Thus for the LABORATORY component, I propose going one step at a time, and focusing on FTL effects,
in effect, for openers. Again, one or two colossal shocks per paper is enough...)

Thus... I guess we cannot avoid SOME degree of complexity here. We probably need to discuss BOTH
<f|exp(iHt)|i> AND time-symmetric objects, to justify your cosmological idea.

AND FURTHERMORE: I don't think that direct heat-based telescopes are so practical.

Maybe. Maybe not.

The thermal aspects here are much trickier than the optical-frequency issues. Maybe we should discuss both, in the end.
We could propose TESTING ideas about the backwards-time temperature-based effects by playing with
high-temperature absorber effects in SPDC experiments. **IF** such effects work, we should be able to
see them even in the laboratory. I guess they probably should work -- how sure are we? But even
if they don't work, SPDC at optical frequencies should work. BUT IN ANY CASE, we can't hope
to design or predict such tricky experiments without some kind of mathematical model to fall back on.
The usual basic QFT model, based on the most-highly-standard S-matrix formalism (and associated
density matrix formalisms and quantum trajectory simulation (QTS) modeling systems) would all
predict ZERO effect.

By the way... leaders in quantum optics experimentation tell me that QTS packages are THE workhorse today
in predicting experiments. QTS would predict that ALL of these experiments would yield zilch results.
Again, we need to be suggesting modifications to QTS -- small enough to still fit today's known results,
but large enough to predict something new.

--------
If there are other ways, besides a combination of SPDC and high-temperature sources, to provide laboratory tests
proving a need to modify QTS... how practical are they, do you think? Our goal here is to
spawn the ACTUAL implementation of ACTUAL experiments, that change people's minds, ala Michelson and Morley.
It comes down to how worked out it is, what we personally have access to, what it costs, who funds it, and
where to find the lab personnel. **IF** our joint paper is well enough accepted by even 20 percent of the community, those problems
may be MUCH easier to address -- but only if we help OTHER people address these practical questions. This is
one reason why I am grateful for the interest of Yanhua Shih, who has gone through the process of physically
setting up paradigm-busting experiments that really convinced the mainstream (in all but one case still on the edge of heresy...).

If there is a thermal-bath experiment that would help, somehow, that's great... though I wonder who would do it, and what
the numbers would look like...

It would be OK to include it in a list of possible things to try, if it is worked out up to a certain point... so long as we
don't oversell the hope... but I'm glad it is not the only option on the list.
But I hope there can be more details in the future which raise the proabbility of its success...

There's an example of the use of this general technique is section 7 of the paper you'll find at:

http://www.usyd.edu.au/philosophy/price/preprints/history.html

This paper covers similar ground to ch 5 of my book, but does it better.