>1) C40 and RFC3 form a series resonant circuit at the input of the
>receiver. My calculations show that it is resonant at 7.34 Mhz. This is
>outside the 40 meter cw band. How does this effect receiver performance?
>
But did you calculate the Q? Very low. There's about 50 ohms
at the antenna-side. And on the receiver side, at LEAST 50 ohms more.
So Q is MAXIMUM 5. That covers the whole band.
And if there IS any left-over reactance, T1 can tune some of it
out.

>2) The diodes D7-D10 are meant to limit input signal. Why are they in
>the middle of the series resonant circuit made up of C40 and RFC3?
>
If the diodes weren't there, the voltage at this point would be
Q6's collector voltage (25v p-p) MULTIPLIED by the Q calculated above (5).
And U1 (or the RFgain pot) would certainly get fried by all that QRP power.
Not only do the diodes clip it down to about 2.6vp-p, but they destroy
the Q of the series-resonant circuit of C40-RFC3, since while conducting,
they're very low-Z devices. So while transmitting, RFC3 is acting as a
proper choke, with 440 ohms inductive reactance in series with the receiver.
But while receiving, this reactance is tuned out by C40, because the
diodes are NOT conducting.

>3) Explain what is happening impedance-wise with the RF gain pot, T1 and
>U1. Why is C1 150 pf instead of 47 pf like in the transmit mixer?
>
C1 is a goof. It IS 47pf.
T1 is tuned to 7Mhz. Don't know what the turns-ratio is, but i'll bet it
steps up the 50 ohm antenna Z to something higher for U1's input of 1500
ohms.

>4) Why is the signal for U1 taken from the center tap of T1, not across
>the whole thing? I thought we would need as much signal as possible from
>the antenna to make it to the receive mixer.
>
I intend to find the turns ratio of T1 (don't know it yet). The
tap could be anywhere, not necessarily at center. Using the tap instead
of the top means the loaded Q of T1 is high. These canned tuned transformers
usually have unloaded Q's approaching 100. With only ONE tuned circuit at
the front end, we need a high Q to reject spurs, and images. Tapping U1
down maintains the Q reasonably high.
Its also a good idea to tap-down too low rather than too high.
This'll improve U1's large-signal handling.

Just measured its turns ratio (roughly):
If you take the full tuned winding length as 100%: