learn - yeah me too!
if you convert the spurs to absolute power, relative to the carrier power - you can add them all up and see they are quite small. power in dBm is 10 log (power in mw). and convert something like a spur that is 20 dB down to a ratio of 0.01 (factor of 100). so a 1 watt signal with 2 spurs 20 dB down, the power in the spurs is something like 0.02 watts !
but keep asking yourself and others questions - and not latching onto your first thought on what may be causing an issue [don't research the web to see how I have resembled this in the past on several projects!]
thanks for what to look for. I never did build that RF probe or PHSNA yet (I will get the built someday and may not need them then ....) -- but I can do a simple sweep with my power meter and dummy load to see what is going on (I would be doing it except for being at work). it seems that issue is after (on transmit) the two sets of 45 MHz filters it now has. [yeah I cracked 2 SMT caps making this mod - but they were tombstone installed - so I realize they are delicate - I may have used leaded ones if my friend who kitted the filter told me the cap value ... our spur fix is this xtal filter (with 2 matching transformers) + Gordon's board of relays - we will have data on 23 units some day]
other than the response of the LPF, the only other thing I see that offers potentially narrow bandwidth are those chokes feeding the PA transistors - and that seems highly unlikely unless a hole bunch of turns were added to it.
out of the box my ubitx was weaker on 40m - and I just brought it down further. its nice on every other band. [yes I blamed it on the xtal filter before I checked all the other bands. ditto for snow and ice on my elevated vertical coil -- always check with a 50 ohm dummy load, or a simpler antenna not impact by ice's dielectric loading!]