I do not know what you are attempting to measure. HF? Out of band transmission, harmonics etc...
Might help if you explained what you are attempting to measure. Frequency range might be helpful.
For HF, up to 30 MHz, the HP 3746 Selective Level Meter does an excellent job. Will give you Freq and Power in a narrow range.
In my experience it is very accurate. Freq readout accuracy is of course limited by the oven oscillator, which is really good, if you set it carefully.
This is old, heavy, big used equipment back from telephone line days. Takes a while to learn how to use it. I picked up one really cheap. Not a lot of demand. Probably cost 10K or more new.
Some telephone lines carried signals in channels throughout the HF spectrum up to 30 MHz and this thing can analyze pretty much anything in those frequencies.
But I do not know if you are in HF or VHF or what?
A lot of Spectrum Analyzers will give you a pretty good power level for each peak and, in general this is very useful and more than adequate for most applications. Even the cheap ones can do a decent job these days. And they are good instruments for all sorts of measurements. I really think this is probably the was to go.
S meters on radios are not accurate but are much better than nothing. I do not know what the standard is for your SDR but even though S meters are supposed to be "calibrated" they are very loose in their readings. One rig will not read the same as another brand and signal strength reading on the S meter will vary from band to band. Maybe the new SDR radio S meters are better, I dunno.
Radiolistener seems to have done this with good results. requires calibration with an input standard, probably at the frequency of interest.
S9 is supposed to be −73 dBm 50.12 pW