C1 and R2 are a highpass filter which should send noise in the upper 100MHz against ground. - Can I ignore this part for the calculation?

No, you cannot ignore this.
Compared to the other resistors, R2 is almost a short circuit. So, C1 is parallel to C2 (next to C5 already mentioned by X J).

I assume that R1 is my load resistor and I can ignore the R2-R5 for loading.

No, you neither can ignore those resistors. Compared to R1 you can ignore R2: you can approximate it as a short circuit as described above.
You can definitely not ignore R3+R4 and R5 because they have (about) the same value as R1.

Conclusion:
τ is defined by C2 and C5 as well as C1 and by R1 as well as R3+R4 and R5.
As X J already pointed out: C2 is hardly detectable as it is even smaller than the tolerances of C5, and as shown above, of C1.

C1 and R2 are a highpass filter which should send noise in the upper 100MHz against ground. - Can I ignore this part for the calculation?

No, you cannot ignore this.
Compared to the other resistors, R2 is almost a short circuit. So, C1 is parallel to C2 (next to C5 already mentioned by X J).

I assume that R1 is my load resistor and I can ignore the R2-R5 for loading.

No, you neither can ignore those resistors. Compared to R1 you can ignore R2: you can approximate it as a short circuit as described above.
You can definitely not ignore R3+R4 and R5 because they have (about) the same value as R1.

Conclusion:
τ is defined by C2 and C5 as well as C1 and by R1 as well as R3+R4 and R5.

As X J already pointed out: C2 is hardly detectable as it is even smaller than the tolerances of C5, and as shown above, of C1.

Next to the tolerances, capacitor value can change with temperature and applied voltage as well. Measuring with an RCL meter will not cancel these last parameters out.